Elif Fatma Sen BW.Indd

Home , ATC code J, ATC code P03, Benzonatate, Clobutinol, Eprazinone, Eprozinol

Use and Safety of Respiratory Medicines in Children

E. F. Şen

EEliflif FatmaFatma SSenen BW.inddBW.indd 1 003-01-113-01-11 15:1715:17 The work presented in this thesis was conducted at the Department of Medical Informatics of the Erasmus University Medical Center, Rotterdam.

The research reported in thesis was funded by the European Community’s 6th Framework Programme. Project number LSHB-CT-2005-005216: TEDDY: Task force in Europe for Drug Development for the Young.

The contributions of the participating primary care physicians in the IPCI, Pedianet and IMS-DA project are greatly acknowledged.

Financial support for printing this thesis was kindly provided by the department of Medical Informatics – Integrated Primary Care Information (IPCI) project of the Erasmus University Medi- cal Center; and by the J.E. Jurriaanse Stichting in Rotterdam.

Cover: Optima Grafi sche Communicatie Printed by: Optima Grafi sche Communicatie

Elif Fatma Şen Use and Safety of Respiratory medicines in Children ISBN: 978-94-6169-003-6 © E.F. Şen, Rotterdam, the Netherlands, 2011. All rights reserved. No part of this thesis may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, without prior written permission of the holder of the copyright.

EEliflif FatmaFatma SSenen BW.inddBW.indd 2 003-01-113-01-11 15:1715:17 Use and Safety of Respiratory Medicines in Children

Het gebruik en de bijwerkingen van respiratoire medicijnen in kinderen

Proefschrift

Ter verkrijging van de graad van doctor aan de Erasmus Universiteit Rotterdam op gezag van de rector magnifi cus

Prof.dr. H.G. Schmidt

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op woensdag 9 februari 2011 om 13.30 uur

door

Elif Fatma Şen

Geboren te ‘s Gravenhage

EEliflif FatmaFatma SSenen BW.inddBW.indd 3 003-01-113-01-11 15:1715:17 Promotiecommissie

Promotor: Prof.dr. M.C.J.M. Sturkenboom

Overige leden: Prof. dr. A. Ceci Prof. dr. J.C. de Jongste Prof.dr. J.N. van den Anker

Co-promotor: Dr. K.M.C. Verhamme

EEliflif FatmaFatma SSenen BW.inddBW.indd 4 003-01-113-01-11 15:1715:17 Contents

Chapter 1 General introduction 7

Chapter 2 Drug utilization 15

2.1 The TEDDY Network: Trends in pediatric drug use in Europe, 17 an epidemiological perspective

2.2 Drug use in children: cohort study in three European countries 25

2.3 Assessment of Pediatric asthma drug use in three European 51 countries; a TEDDY study

2.4 Eff ects of safety warnings on prescription rates of cough and cold 73 medicines in children below 2 years of age

Chapter 3 Drug safety 89

3.1 Asthma medicines related adverse drug reactions in children: 91 Results from the WHO adverse events database

3.2 Cough and cold medicines related adverse drug reactions in 105 children: Results from the WHO adverse events database

3.3 Use of steroids in children and risk of fractures: a nested case- 123 control study

3.4 Steroid induced pancreatitis in children: results from a population- 137 based case-control study

Chapter 4 General discussion 149

Summary 167

Samenvatting 173

PhD portfolio 179

Dankwoord 183

List of publications 187

About the author 191

EEliflif FatmaFatma SSenen BW.inddBW.indd 5 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 6 003-01-113-01-11 15:1715:17 Chapter 1

General introduction

EEliflif FatmaFatma SSenen BW.inddBW.indd 7 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 8 003-01-113-01-11 15:1715:17 Chapter 1: General introduction

The lack of appropriately authorised and formulated medicines for use in the pediatric population is a longstanding problem and cause for concern. As a result, most medicines are prescribed to children on an off -label or an unlicensed basis [1]. Dosing regimens approved for adults are extrapolated to pediatric age groups, for example on the basis of proportionality of weight, without pediatric pharma- cokinetic or pharmacodynamic data. Safety and effi cacy are presumed to be the same as in adults, but this may not be the case. Evidence-based information in children is often not readily available, and prescribing decisions may have to be based on accepted practice presented in formularies. On January 26, 2007, the new regulation of the European Union (EU) on medicinal products for pediatric use came into force [2]. The overall aim of the regulation is to improve the health of children in the EU. More specifi cally the objectives include: increasing the development of medicines for use in children; ensuring that medicines used to treat children are subject to high quality research and appropriately authorised for use in children; improving the information available on the use of medicines in children; and achieving the above objectives without subjecting children to unnecessary clinical trials or delaying the authorisation of medicines in the adult population. Previously, pharmaceutical industry had been free to restrict a particular medicine’s development to the adult population; the development of pediatric medicines was solely at the industry’s discretion. The new regulation requires that all applications for marketing authorisation for new medicines, must contain the results of all studies and information required in a previously agreed pediatric investigation plan (PIP) The PIP will contain a full proposal of all the studies (and their timings) necessary to support the pediatric use of an individual product and will cover all pediatric age groups and all necessary age-appropriate formulations. In some case, pharmaceutical companies can obtain a deferral of waiver from a PIP, for example where there is no pediatric therapeutic need or where pediatric use of the product is not appropriate. Unless a deferral or a waiver has been granted, pediatric data must be provided in all applications for the authorisation of new medicinal products. The regulation also establishes a new type of marketing authorisation, called the pediatric use marketing authorization (PUMA), intended to stimulate the development of off -patent products for use in the pediatric population. Only medicines that are intended solely for use in children will be eligible for a PUMA. The new legislation obliges the European Medicines Agency to build one European network from the diff erent existing networks and collaborative projects, investigators, and specifi c Centers of Excellence.

TEDDY The Task-force in Europe for Drug Development for the Young (TEDDY) Network of Excellence was established in 2005 through funding of the European Community’s sixth Framework

EEliflif FatmaFatma SSenen BW.inddBW.indd 9 003-01-113-01-11 15:1715:17 Programme to contribute to the promotion of safe and effi cacious medicines for children in the context of the impending European Paediatric Regulation that came into force in January 2007 [3, 4]. The overall aim of the TEDDY NoE is to promote the availability of safe and eff ective medicines for children in Europe by integrating existing expertise and good practices, as well as stimulating pediatric drug development. The work presented in this thesis is a spin-off of the TEDDY activities.

Respiratory medicines Respiratory medicines contain a wide variety of medicines and are amongst the most frequently used medicines in children [5]. In this thesis we will focus on the use and safety of two groups of respiratory medicines in particular, namely asthma medicines and cough and cold medicines.

Asthma medicines Asthma is defi ned as a chronic infl ammatory disorder of the airways and is associated with airway hyperresponsiveness that leads to recurrent episodes of wheezing, breathlessness, chest tightness, and coughing [6]. Asthma is the most common chronic disease of childhood and the leading cause of childhood morbidity from chronic disease as measured by school absences, emergency department visits and hospitalizations [7]. As a result, asthma medicines are used widely in the pediatric population. According to the Global Initiative for Asthma (GINA), bronchodilators (inhaled short-acting ß2-mimetics: SABA; long-acting ß2-mimetics: LABA; anticholinergics: ACH), inhaled corticosteroids (ICS) and alternative (add-on) treatments such as leukotriene receptor antagonists (LTRA), xanthines and anti-allergics (cromones) all have a place in the treatment of childhood asthma, although at various stages depending on severity of illness [6]. Apart from these respiratory drugs, systemic glucocorticoids are also used in the treatment of asthma. However, due to side eff ects such as potential adrenal suppression and growth impairment with long term use [6, 8], use of systemic glucocorticoids in asthmatic children is restricted to asthma exacerbations and very severe persistent asthma [6]. Despite the widespread use of these medicines by children very little is know about the long term safety. The Pediatric Committee of the European Medicines Agency has recently published a pediatric needs list, containing drugs that are being used in children and where information on pharmacokinetics, effi cacy and safety is urgently needed [9]. This list includes asthma medicines such as salbutamol, fl uticasone, montelukast and many others. The missing information on the pediatric needs list is mostly addressed by through performing clinical trials in children, aimed to obtain a PUMA. However, a TEDDY study has shown that a lot of data on pediatric drug usage is readily available in many healthcare databases throughout Europe [10]. These databases provide an enormous potential to conduct pediatric pharmacoepidemiological research, without subjecting children to unnecessary clinical trials. In this thesis, we will demonstrate that data on use and safety of asthma medicines in pediatrics can be obtained by using healthcare databases which are readily available.

EEliflif FatmaFatma SSenen BW.inddBW.indd 1100 003-01-113-01-11 15:1715:17 Chapter 1: General introduction

Cough and cold medicines Cough and cold medicines (CCMs) are frequently used to treat upper respiratory tract symptoms in children and have been on the market for many decades. This group of medicines includes expectorants, mucolytics, opium alkaloids (for cough), nasal sympathicomimetics and anti-histamines. In the past years, safety issues have been raised about the use of these drugs in young children. In the United States poison-control centers have reported more than 750,000 calls of concern related to cough and cold medicines since January 2000 [11]. A report from the Centers for Disease Control and Prevention identifi ed more than 1500 emergency room visits in 2004 and 2005 for children under 2 years of age who had been given cough or cold medicines [12]. A review by the Food and Drug Administration (FDA) identifi ed 123 deaths related to the use of such products in children under six during the past several decades [13]. A growing number of studies associated the use of CCMs to serious adverse events in children, such as cardiac arrhythmias, depressed levels of consciousness and even death [14]. In 2007, a citizen petition was sent to the FDA, urging the agency to review the eff ects of cough and cold medicines in young children. As a result, the FDA released a recommendation that cough and cold medicines should not be used in children below two years of age [15]. Con- sequently, Canada and the United Kindgom also posted warnings against the use of CCMs in young children [16, 17]. In this thesis, we studied the eff ects of these regulatory actions on prescription rates of CCMs. We also investigated the type of adverse eff ects related to the use of CCMs in children that were reported to the Vigibase database on spontaneous reporting.

Data sources used in this thesis Although randomized controlled trials (RCTs) are the gold standard to assess the effi cacy of a drug, observational studies, using data from large health care databases are ideal to assess the utilization and safety of a drug. [18]. As previously mentioned, a TEDDY study showed that many European healthcare databases are readily available, providing an enormous potential for pediatric pharmacoepidemiological research [10]. The conclusion of that study was that future research should focus on methods to bring data from diff erent databases together to use the full capacity eff ectively. In this thesis, we combined data from some of these databases to obtain data on utilization and safety of asthma medicines and cough cold medicines. Three diff erent types of data sources were used for the studies presented in this thesis: primary care physician databases (IPCI, Pedianet and IMS), the PHARMO Record Linkage System, and the Vigibase database of the WHO-UMC which collects spontaneous reports on adverse drug reactions. These data sources allowed us to perform large scale pharmacoepidemiologic and pharmaco- vigilance studies, without limitations mostly observed in RCTs, such as small sample sizes and

EEliflif FatmaFatma SSenen BW.inddBW.indd 1111 003-01-113-01-11 15:1715:17 short duration of follow-up [19]. We demonstrate that an enormous amount of healthcare data is available in databases throughout Europe and that these databases can be used to provide valuable information on use and safety of medicines in children.

Aims and outline of this thesis In this thesis we aimed to analyse the use and safety of asthma medicines and cough and cold medicines in children through large scale observational studies, by combining readily available healthcare databases. In chapter 2 we describe the utilization of medicines in children and in particular, the utilization of asthma and cough and cold medicines. For these studies we used the IPCI, Pedianet and IMS-DA databases In chapter 3 safety studies are presented. To get a broader understanding of the adverse drug reactions causes by asthma medicines and CCMs, two studies focus on adverse drug reactions for reported to Vigibase. In two other studies we assess the association between steroids (both inhaled and oral) and two outcomes: fractures and pancreatitis, by means of case control studies. For these studies we used the IPCI and Pedianet databse and PHARMO RLS. The meaning and limitations of the studies are discussed in chapter 4 and recommendations are made for future research.

EEliflif FatmaFatma SSenen BW.inddBW.indd 1122 003-01-113-01-11 15:1715:17 Chapter 1: General introduction

References

1 Conroy S, Choonara I, Impicciatore P, Mohn A, Arnell H, Rane A, et al. Survey of unlicensed and off label drug use in paediatric wards in European countries. European Network for Drug Investigation in Children. BMJ (Clinical research ed. 2000 Jan 8;320(7227):79-82. 2 Dunne J. The European Regulation on medicines for paediatric use. Paediatric respiratory reviews. 2007 Jun;8(2):177-83. 3 Regulation (EC) No. 1901/2006 of the European Parliament and of the Council of 12 December 2006 on medicinal products for paediatric use and amending Regulation (EEC) No. 1768/92, Directive 2001/20/EC, Directive 2001/83/EC and Regulation (EC) No. 726/2004 (Text with EEA relevance). [cited; Available from: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2006:378:0001:0019:EN: PDF 4 The Task-force in Europe for Drug Development for the Young. [cited; Available from: http://www. teddyoung.org/ 5 Sturkenboom MC, Verhamme KM, Nicolosi A, Murray ML, Neubert A, Caudri D, et al. Drug use in children: cohort study in three European countries. BMJ (Clinical research ed. 2008;337:a2245. 6 Risselada R, Straatman H, van Kooten F, Dippel DW, van der Lugt A, Niessen WJ, et al. Withdrawal of statins and risk of subarachnoid hemorrhage. Stroke. 2009 Aug;40(8):2887-92. 7 Masoli M, Fabian D, Holt S, Beasley R. The global burden of asthma: executive summary of the GINA Dissemination Committee report. Allergy. 2004 May;59(5):469-78. 8 Szefl er SJ. Glucocorticoid therapy for asthma: clinical pharmacology. The Journal of allergy and clinical immunology. 1991 Aug;88(2):147-65. 9 Trifi ro G, Fourrier-Reglat A, Sturkenboom MC, Diaz Acedo C, Van Der Lei J. The EU-ADR Project: Pre- liminary Results and Perspective. Stud Health Technol Inform. 2009;148:43-9. 10 Neubert A, Sturkenboom MC, Murray ML, Verhamme KM, Nicolosi A, Giaquinto C, et al. Databases for pediatric medicine research in Europe--assessment and critical appraisal. Pharmacoepidemiology and drug safety. 2008 Dec;17(12):1155-67. 11 Consumer Healthcare Products Association testimony before the Food and Drug Administration (October 18, 2007). Washington, DC: American Association of Poison Control Centers, 2007. 12 Infant deaths associated with cough and cold medications--two states, 2005. Mmwr. 2007 Jan 12;56(1):1-4. 13 Food and Drug Administration, Division of Drug Risk Evaluation. Nonprescription Drug Advisory Committee meeting: cold, cough, allergy, bronchodilator, antiasthmatic drug products for over-the- counter human use. 2007:29. memorandum. [cited 8/11/2010]; Available from: http://www.fda.gov/ ohrms/dockets/ac/07/briefi ng/2007-4323b1-02-FDA.pdf 14 Sharfstein JM, North M, Serwint JR. Over the counter but no longer under the radar--pediatric cough and cold medications. The New England journal of medicine. 2007 Dec 6;357(23):2321-4. 15 FDA. FDA Recommends that Over-the-Counter (OTC) Cough and Cold Products not be used for Infants and Children under 2 Years of Age. 2008 [cited 03-10-2009]; Available from: http://www.fda. gov/Drugs/DrugSafety/PublicHealthAdvisories/ucm051137.html 16 Canada H. Health Canada Advisory: Cough and Cold products for children. 2009 [cited 07-10-2009]; Available from: http://chd.region.waterloo.on.ca/web/health.nsf/vwSiteMap/E7742A6975D6895E85 2570E50055EEFF/$fi le/PHYSUP_JAN09.pdf?openelement 17 MHRA. Press release: Better medicines for children’s coughs and colds. 2009 [cited 03-10-2009]; Avail- able from: http://www.mhra.gov.uk/NewsCentre/Pressreleases/CON038902

EEliflif FatmaFatma SSenen BW.inddBW.indd 1133 003-01-113-01-11 15:1715:17 18 Sorensen HT, Lash TL, Rothman KJ. Beyond randomized controlled trials: a critical comparison of trials with nonrandomized studies. Hepatology (Baltimore, Md. 2006 Nov;44(5):1075-82. 19 Morales-Olivas FJ, Morales-Carpi C. Clinical trials in children. Reviews on recent clinical trials. 2006 Sep;1(3):251-8.

EEliflif FatmaFatma SSenen BW.inddBW.indd 1144 003-01-113-01-11 15:1715:17 Chapter 2

Drug utilization

EEliflif FatmaFatma SSenen BW.inddBW.indd 1155 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 1166 003-01-113-01-11 15:1715:17 Chapter 2.1

The TEDDY network: Trends in pediatric drug use in Europe, an epidemiological perspective

Elif Fatma Sen, Katia MC Verhamme, Ron Herings, Antje Neubert, Mariagrazia Felisi, Adriana Ceci, Miriam CJM Sturkenboom.

Published in: European Journal of Hospital Pharmacy Practice, 2007; Nr: 6; Pages: 22- 24;

EEliflif FatmaFatma SSenen BW.inddBW.indd 1177 003-01-113-01-11 15:1715:17 Summary

Not only do we have limited drug effi cacy data for children, but also little is known about the extent, types and safety of drugs that are actually being used in pediatrics. The TEDDY network aims to improve pediatric drug development. To help in prioritizing research the network has started out to describe available datasources for pharmacoepidemiologic research and the actual description of drug utilization in children. This paper describes these TEDYY activities in more detail.

EEliflif FatmaFatma SSenen BW.inddBW.indd 1188 003-01-113-01-11 15:1715:17 Chapter 2.1: The TEDDY network

Introduction

The Taskforce in Europe for Drug Development for the Young (TEDDY) is a Network of Excel- lence funded under the 6th EU Framework (www.teddyoung.org). The overall aim of TEDDY is to promote the availability of safe and eff ective medicines for children in Europe by integrating existing expertise and good practices, as well as stimulating pediatric drug development. One of the objectives of TEDDY is to set up a harmonised, integrated and reliable European database containing information on the types and extent of drugs used in children. This paper reviews activities related to assessing pediatric drug use. The European Medicines Evaluation Agency (EMEA) has drafted lists of drugs in need of pediatric research, however with this list of more than 100 drugs it is necessary to prioritize, since it will be impossible to conduct all requested research in parallel (reference necessary!! EMEA site). Using data on which types of drugs are actually used in children may help in choosing for which drugs we should most urgently demonstrate effi cacy in children, fi nd adequate formulations and provide data on short and long term safety.

Datasources

Information on drug use in general can be obtained from market research sources, but these do not provide information on age, or other clinical details such as the type of prescriber, indication, dose and off -label use. Pharmacoepidemiology (the study of the utilization and eff ects of drugs in large populations) relies heavily on health care databases as data source for adult drug use and eff ects. To understand better whether databases would be suitable for pediatric postmarketing drug research a survey was conducted within the TEDDY NoE to describe the characteristics of existing European healthcare databases and to assess whether the databases can be used for pediatric medicines research [1]. Siventeen healthcare databases from 10 diff erent European countries were identifi ed, together covering a source population of about 9 million children from 0-18 years of age. Almost exclusively these databases covered only out-patient drug use, Ten out of the 17 databases previously published studies specifi cally addressing the paediatric population, but only three databases (PEDIANET, GPRD, IPIC) were already used to investigate the safety of paediatric drugs. [2-9]. An international collaboration or network of databases as foreseen in the TEDDY project would allow for the extraction of important information from data already available and enhance our knowledge about pediatric drug use and safety of e.g. off -label used drugs.

EEliflif FatmaFatma SSenen BW.inddBW.indd 1199 003-01-113-01-11 15:1715:17 Outpatient drug use

To demonstrate the possibilities of such an international collaboration and to provide information on pediatric drug use in Europe, TEDDY started describing drug prescription patterns in primary care by using established medical record databases in the UK, Netherlands and Italy. These databases contain information on more than fi ve million paediatric prescriptions during 2.3 million pediatric person years). The prevalence of drug use (number of users per year) could be divided in three categories of use namely high use (>10% of children using the drug at least once in a year), moderate use (1-10% users per year) and low use (<1% of users per year). For all age categories, anti-infectives (ATC J), dermatological (ATC D) and respiratory drugs (ATC R) were in the high use group, whereas cardiovascular and anti-neoplastic drugs were always in the low use group[10] (fi gure 1).

<2 yrs. 2-11 yrs. 12-18 yrs

50.00 45.00 40.00 35.00 30.00 25.00 year 20.00 15.00 10.00 5.00

% of children using drug in using drug children % one of 0.00 ABCDGHJ LM NPRS Anatomic drug class (ATC) Figure 1: Pediatric drug use in Netherlands, UK and Italy, by age and anatomic class Legend: ATC A: Alimentary tract, ATC B: Blood and bloodforming organs, ATC C: cardiovascular, ATC D: Dermatological, ATC G: Genitourinary tract, ATC H: Hormones (excl sex), ATC J: Anti-infectives, ATC L: Anti-neoplastic and immunological, ATC M: Musculoskeletal, ATC N: Nervous system, ATC P: Anti-parasitic drugs, ATC R: Respiratory, ATC S: Sensory organs, ATC V: various.

Since the primary care database may miss prescriptions written by specialists we also searched for databases that could provide insight in drug use by type of prescriber. The PHARMO record linkage database in the Netherlands comprises pharmacy dispensing data on more than 360.000 children (www.pharmo.nl). A study on all drugs dispensed in the year 2005 showed that the majority of pediatric drug dispensings (78%) are prescribed by the general practitioners (fi gure 2). The high volume categories dermatological, respiratory and anti-infective drugs are mostly prescribed by GPs (all less than 15% by specialists). Categories with a relative high percentage of prescribing by specialists are the alimentary tract drugs (ATC A: 49% by specialists), drugs for blood and blood forming organs (ATC B: 46% by specialists), hormones (ATC H: 40% by specialists), antineoplastic and immunomodulating drugs (ATC L: and neuropsychiatric

EEliflif FatmaFatma SSenen BW.inddBW.indd 2020 003-01-113-01-11 15:1715:17 Chapter 2.1: The TEDDY network

specialist GP

25%

20%

15%

10% % of %dispensing of 5%

0% ABCDGH J LMNP RSV ATC anatomic level Figure 2: Percentages of pediatric drug dispensings written by specialists or GPs (percentages calculated on the basis of the total number of drug prescriptions) Legend: ATC A: alimentary tract, ATC B: Blood and bloodforming organs, ATC C: cardiovascular, ATC D: dermatological, ATC G: genitourinary, ATC H: hormones, ATC J: anti-infectives, ATC L: anti-neoplastic and immunological, ATC M: musculoskeletal, ATC N: nervous system, ATC P: anti-parasitic drugs, ATC R: respiratory, ATC S: sensory organs, ATC V: various.

drugs (ATC N: 47%). Within the alimentary drug class, the stomatological preparations group (A01) was responsible for most of the specialist prescriptions, within the ATC B class specialists prescribed mostly antihemorrhagics (B01) and antianemic (B02) drugs, within the class of hormones oral steroids (H02) and thyroid therapy (H03) were mostly prescribed by specialists, within the antineoplastic and immunomodulating class all four groups (antineoplastic drugs, endocrine therapy, immunostimulants and immunosuppressive drugs) were prescribed mostly by specialists, within the neuropsychiatric drugs, antiepileptics (N03), psycholeptics (N05) and psychoanaleptics (N06) were mostly prescribed by specialists (fi gure 3). Prescription data from general practices is a reliable source to assess drug exposure in countries where the general practitioner acts as gatekeeper to further care, dispensing data (including outpatient specialists) is available as claims data in countries with a national health system (e.g. Scandinavia, and Italy) but frequently are government owned and more diffi cult to access. Apart from some ad hoc studies very little is known about pediatric drug use in the hospital [11], whereas it has been demonstrated that most of the inpatient pediatric drugs are used off -label [12, 13]. There is an urgent need for collaboration between pediatric hospital pharma- cies and paediatric researchers to perform quantitative and qualitative studies on this subject. TEDDY is currently looking for partners that can participate in such studies.

EEliflif FatmaFatma SSenen BW.inddBW.indd 2121 003-01-113-01-11 15:1715:17 GP specialist

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% A01 A02 A03 A04 A05 A06 A07 A08 A09 A10 A11 A12 A16 L01 L02 L03 L04 B01 B02 B03 B05 H01 H02 H03 H04 H05 N01 N02 N03 N04 N05 N06 N07 ABHLN % of dispensing within anatomic level Figure 3: Percentage of pediatric dispensed drug prescriptions written by specialists and GPs, as percentage of total number of prescriptions within anatomic level Legend: ATC A01: Stomatological preparations, A02: Drugs for acid related disorders, A03: Drugs for functional gastrointestinal disorders, A04: Antiemetics and antinauseants, A05: Bile and liver therapy, A06: Laxatives, A07: Antidiarrheals, intestinal anti-infl ammatory/anti-infective agents, A08: Antiobesity preparations, excluding diet products, A09: Digestives, including enzymes, A10: Drugs used in diabetes, A11: Vitamins, A12: Mineral supplements, A16: Other alimentary tract and metabolism products, B01: Antithrombotic agents, B02: Antihemorrhagics, B03: Antianemic preparations, B05: Blood substitutes and perfusion solutions, H01: Pituitary and hypothalamic hormones and analogues, H02: Corticosteroids for systemic use, H03: Thyroid therapy, H04: Pancreatic hormones, H05: Calcium homeostasis, L01: Antineoplastic agents, L02: Endocrine therapy, L03: Immunostimulants, L04 Immunosuppressive agents, N01: Anesthetics, N02: Analgesics, N03: Antiepileptics, N04: Anti-parkinson drugs, N05: Psycholeptics, N06: Psychoanaleptics, N07: Other nervous system drugs.

Unlicensed and off -label

Because of the lack of research on safe and eff ective drugs in children, unlicensed and off -label drug use in children is widespread [11-14]. Previous investigations have shown that various defi nitions for “off -label and unlicensed drug use” are used, which makes it diffi cult to compare data in children. A Delphi method based survey was recently conducted by TEDDY to better defi ne the concepts off -label and unlicensed drug use in children [15]. The defi nition concluded from the survey has been discussed with the European Agency for the Evaluation of Medicinal Products (EMEA) and is agreed by a panel of European experts in the fi led. Off -label is now defi ned as all uses of a marketed drug that are not included in the Summary of Products Char- acteristics (SPC) with reference to indication, dosage, formulation, routine of administration, etc. This defi nition is adopted by the TEDDY expert group and also endorsed by the EMEA. To quantify the extent of off -label use according to the new defi nitions, TEDDY is investigating the number of off -label users and percentage of off -label drug use for the drugs prescribed in

EEliflif FatmaFatma SSenen BW.inddBW.indd 2222 003-01-113-01-11 15:1715:17 Chapter 2.1: The TEDDY network

primary care in the Netherlands, Italy and the UK according to the new defi nitions. This scheme can be applied to any type of datasource and demonstrate the areas of greatest pediatric research needs.

Conclusion

To better guide prioritization of pediatric research it is useful to have information about the extent to which various drugs are actually being used in children in both primary care as well as in secondary and tertiary care. Observational databases consisting of drug prescriptions or dispensings are a good resource to describe which specifi c types and formulations of drugs are actually being used in outpatient settings, TEDDY is providing important information on resources, and outpatient drug use in children However, resources on in-hospital drug use is lacking and collaboration is being sought to create an international network on in-hospital drug databases. Pharmacists being aware of patient level databases of drug use and diagnoses in their hospital can contact the authors to collaborate in creating a pan-European network of in-hospital data for pediatric drug use.

EEliflif FatmaFatma SSenen BW.inddBW.indd 2233 003-01-113-01-11 15:1715:17 References

1 Antje Neubert MCS, Macey L. Murray, Katia MC Verhamme, Alfredo Nicolosi, Carlo Giaquinto, Adriana Ceci, Ian C. Wong. Databases for Pediatric Medicine Research in Europe - Assessment and Critical Appraisal. submitted. 2007. 2 Andrews N, Miller E, Grant A, Stowe J, Osborne V, Taylor B. Thimerosal exposure in infants and developmental disorders: a retrospective cohort study in the United kingdom does not support a causal association. Pediatrics. 2004 Sep;114(3):584-91. 3 Hypponen E, Laara E, Reunanen A, Jarvelin MR, Virtanen SM. Intake of vitamin D and risk of type 1 diabetes: a birth-cohort study. Lancet. 2001 Nov 3;358(9292):1500-3. 4 Kilkkinen A, Virtanen SM, Klaukka T, Kenward MG, Salkinoja-Salonen M, Gissler M, et al. Use of antimi- crobials and risk of type 1 diabetes in a population-based mother-child cohort. Diabetologia. 2006 Jan;49(1):66-70. 5 Schlienger RG, Jick SS, Meier CR. Inhaled corticosteroids and the risk of fractures in children and adolescents. Pediatrics. 2004 Aug;114(2):469-73. 6 Smeeth L, Hall AJ, Fombonne E, Rodrigues LC, Huang X, Smith PG. A case-control study of autism and mumps-measles-rubella vaccination using the general practice research database: design and methodology. BMC public health. 2001;1:2. 7 Sturkenboom M, Nicolosi A, Cantarutti L, Mannino S, Picelli G, Scamarcia A, et al. Incidence of muco- cutaneous reactions in children treated with nifl umic acid, other nonsteroidal antiinfl ammatory drugs, or nonopioid analgesics. Pediatrics. 2005 Jul;116(1):e26-33. 8 t Jong GW, Eland IA, Sturkenboom MC, van den Anker JN, Stricker BH. Determinants for drug prescribing to children below the minimum licensed age. European journal of clinical pharmacology. 2003 Feb;58(10):701-5. 9 van Staa TP, Cooper C, Leufkens HG, Bishop N. Children and the risk of fractures caused by oral corticosteroids. J Bone Miner Res. 2003 May;18(5):913-8. 10 Miriam CJM Sturkenboom KMV, Macey L. Murray, Antje Neubert, Ian C. Wong, Daan Caudri, Gino Picelli, Elif Fatma Sen, Carlo Giaquinto, Luigi Cantarutti, Alfredo Nicolosi, Paola Baiardi, Adriana Ceci, on behalf of the TEDDY European Network of Excellence. Drug Utilisation in Children. submitted. 2007. 11 Shah SS, Hall M, Goodman DM, Feuer P, Sharma V, Fargason C, Jr., et al. Off -label drug use in hospitalized children. Archives of pediatrics & adolescent medicine. 2007 Mar;161(3):282-90. 12 t Jong GW, van der Linden PD, Bakker EM, van der Lely N, Eland IA, Stricker BH, et al. Unlicensed and off -label drug use in a paediatric ward of a general hospital in the Netherlands. European journal of clinical pharmacology. 2002 Jul;58(4):293-7. 13 t Jong GW, Vulto AG, de Hoog M, Schimmel KJ, Tibboel D, van den Anker JN. A survey of the use of off -label and unlicensed drugs in a Dutch children’s hospital. Pediatrics. 2001 Nov;108(5):1089-93. 14 t Jong GW, Vulto AG, de Hoog M, Schimmel KJ, Tibboel D, van den Anker JN. Unapproved and off -label use of drugs in a children’s hospital. The New England journal of medicine. 2000 Oct 12;343(15):1125. 15 Antje Neubert P, Alessandro Bonifazi2, Mariana Catapano PhD2, Paola Baiardi, PhD2, Carlo Guiaquinto, MD3 Catherinje Knibbe4 Miriam CJM Sturkenboom, PhD45, Ian CK Wong, PhD1 Adriana Ceci, PhD2 , On behalf of the TEDDY Network of Excellence. Defi ning Off -label and Unlicensed Use of Medicines for Children: Results of a Delphi survey. BJCP (submitted). 2007.

EEliflif FatmaFatma SSenen BW.inddBW.indd 2244 003-01-113-01-11 15:1715:17 Chapter 2.2

Drug utilisation in children -A cohort study in three European countries-

Miriam CJM Sturkenboom, Katia MC Verhamme, Alfredo Nicolosi, Macey L. Murray, Antje Neubert, Ian C. Wong, Daan Caudri, Gino Picelli, Elif Fatma Sen, Carlo Giaquinto, Luigi Cantarutti, Paola Baiardi, Maria- Grazia Felisi, Adriana Ceci on behalf of the TEDDY European Network of Excellence.

Published in: British Medical Journal, Clinical research ed. 2008;337:a2245.

EEliflif FatmaFatma SSenen BW.inddBW.indd 2255 003-01-113-01-11 15:1715:17 Abstract

Context: The new European legislation concerning paediatric drugs has been conceived in order to stimulate research into children’s medicines, but little is known about the patterns of general and chronic drug use in children. Objective: To provide an overview of drug use in children in three European countries. Design: A r etrospective cohort study covering the years 2000-2005 Setting: Three primary care research databases in the Netherlands (IPCI), United Kingdom (IMS- DA) and Italy (Pedianet) Patients: All children up to 14 (Italy) or 18 years of age (United Kingdom and The Netherlands). Main outcome measure. Prevalence of use per year was calculated by drug class (anatomic and therapeutic levels). Within each therapeutic drug level, we estimated the prevalence of “recurrent/chronic” use (>3 prescriptions per year) or “non-recurrent” or “acute” use (<3 prescriptions per year). Within each anatomic class the top fi ve most frequently used drugs are described and evaluated for off -label status. Results: Three levels of drug use could be distinguished in the study population of 675,868 children: high (>10 per year), moderate (1-10 per year) and low (<1 per year). For all age categories, anti-infectives, dermatological and respiratory drugs were in the high use group, whereas cardiovascular and anti-neoplastic drugs were always in the low use group. Emollients, topical steroids and asthma drugs had the highest prevalence of recurrent use, but low prevalence drugs were relatively more often used recurrently than acutely. In the top fi ve of highest prevalence drugs topical inhaled and systemic steroids, oral contraceptives and topical or systemic antifungals were most frequently off -label Conclusion: This paper provides an overview of outpatient pediatric prescription patterns in a large European population. As these data are population-based and cover all types of drugs they may support prioritising research needs concerning long-term paediatric drug safety, as well as areas for need of extra effi cacy and eff ectiveness studies in paediatric medicine.

EEliflif FatmaFatma SSenen BW.inddBW.indd 2266 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries-

BACKGROUND

Recent years have seen growing concerns about the incompleteness of the evidence relating to the effi cacy and safety of the drugs used in children. Almost all of the drugs prescribed to children are the same as those originally developed for adults. They are frequently prescribed to children on an unlicensed or ‘off -label’ basis (percentages ranging from 11-80 )1 simply by extrapolating adult data, without conducting any pediatric clinical, kinetic, dose-fi nding or formulation studies. However, childhood diseases may be very diff erent from their adult equiva- lents, and the processes underlying growth and development may lead to a diff erent eff ect or an adverse drug reaction unseen in adults (Reye’s syndrome is an outstanding example). In order to provide legitimate and appropriate treatment for children’s diseases, new legislation was approved in the US in 2003 and the European Union in 20072. Both the Food and Drug Administration (FDA) and the European Medicines Agency for the Evaluation of Medicinal Products (EMEA) now off er license extensions to marketing authorisation holders who provide evidence concerning the effi cacy and safety in children of new drugs or ‘off -label’ drugs3-6. The World Health Organization underlines the need for these actions and has launched a global campaign in December 2007 to ‘make medicines child size’ with the purpose to address the need for improved availability and access to safe child-specifi c medicines for all children7 In an attempt to assist public authorities and the research community in specifying the research needs, we investigated the use of paediatric drugs in children in three European countries, using population-based primary care prescription data in order to describe the current patterns of drug prescribing for children by primary care physicians (general practitioners or paediatricians).

METHODS

Setting The primary care of children is entrusted to general practitioners in the UK and The Netherlands, and to paediatricians in Italy8 9. Access to health care is free in Italy and the UK, and fully covered by healthcare insurance in The Netherlands. In these countries, primary care physicians are the guardians of children’s health, which means that all clinical information concerning the patients (including summaries of specialist and hospital care) is kept in their medical records. Since all children need to be registered with a GP in NL and UK, and with a family paediatrician in Italy, the databases are population-based9.

Data collection We used the same protocol to study prescription patterns in Italy (IT), The Netherlands (NL) and the United Kingdom (UK), making use of the Pedianet database (paediatric electronic medical

EEliflif FatmaFatma SSenen BW.inddBW.indd 2277 003-01-113-01-11 15:1715:17 records from 150 paediatricians since 2000) in Italy10, the Integrated Primary Care Information (IPCI) database (comprising adult and paediatric electronic medical records from more than 400 doctors since 1996) in The Netherlands8 11 12, and the IMS Disease Analyzer database (IMS- DA: electronic medical records on adults and children from 670 doctors) in the UK13 . All of these databases include the complete automated medical records of primary care, physicians and have been used and proven valid for pharmacoepidemiological research9. The age and gender distribution in the various databases is representative for the country they originate from.

Study population The dynamic study population in each country consisted of all children aged 0-18 years (0-14 years in Italy) who had a database history of at least six months, or who were born during the study period (1 January 2000 – 31 December 2005), and we calculated the person-time of follow-up of each child, stratifi ed by calendar year and age group. Age was assessed on 1 January of each year and grouped according to the Guidelines of the International Conference of Harmonization (ICH) in <2, 2-11, and 12-18 years of age 14 The <2 year age category could not be further stratifi ed into newborns (<1 month) and infants (1-24 months) as exact dates of birth were not available because of privacy regulations. Each child was followed from the start of the study period or the date of registration with the primary practice (whichever was the latest), until the cancellation of the registration with the practice or the end of the study period. The persontime accumulated in each calendar year was used as denominator for calculation of prevalence rates. Over the study period children could contribute to more than one age category.

Drug prescriptions All prescribed drugs in children during follow-up were retrieved from the prescription data in the database. The drug prescriptions were grouped on the basis of the WHO’s Anatomical Therapeutic Chemical (ATC) classifi cation system which made comparison between countries possible.

Statistical analysis User prevalence rates (per 1,000 PY) were estimated by counting the number of children using a specifi c drug in a specifi c calendaryear. The prevalence rates were calculated by age and country to account for diff erences in distributions between populations and to allow for direct within stratum comparisons. User prevalence rates should be interpreted as the number of children per 1,000 children who use a specifi c drug class in one year. Drug use prevalence in Italy was not calculated for children aged 15-18 years because all of the children were censored at the age of 15 years. The use of person-years rather than persons as the denominator was due to the dynamic nature of age and the population.

EEliflif FatmaFatma SSenen BW.inddBW.indd 2288 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries-

For each anatomic drug category we assessed the age and country specifi c user prevalence rates for all individual drugs in 2005. The fi ve drugs with the highest prevalence per anatomic class in each country are reported and were evaluated for off -label status considering age only. A drug was considered to be off -label for age if the child’s age at the time of use, was below the lowest approved age as mentioned in the Summary of Product Characteristics of that drug per country15. Within each therapeutic drug level, we separately estimated the prevalence of children presenting “recurrent/chronic” (>3 prescriptions per year) versus “non-recurrent” or “acute” drug use (<3 prescriptions per year), and the ratio between them in order to help identify the treatments that are more frequently used for chronic than acute paediatric diseases. Comparisons between user prevalence rates were done by means of the Chi-square test.

RESULTS

Study population Our population of 675,868 children generated 2,334,673 person-years of follow-up (Table 1); the mean individual follow-up was 3.5 years per child. Most of the children (65%) came from the IMS database in the UK, 19% from Italy and 16% from The Netherlands. The databases recorded more than fi ve million paediatric prescriptions. The prescription rate was highest for the children aged less than two years in all three countries and, in each age group, was signifi cantly higher in the UK and Italy than in The Netherlands (p-values <0.001),(table 1).

Drug use by anatomically grouped drug classes The highest prevalence rates among the children aged less than two years were for anti- infectives (ATC code J), respiratory drugs (R) and dermatologicals (D), which were used by 48% , 30% and 30% of the children, respectively (Figure 1). The other frequent prescriptions were for gastrointestinal drugs (A, with a user prevalence of 20%), drugs for the nervous system (N, 14%) and drugs for sensory organs (S, 19%). ATC classes B (blood and blood forming organs), H (hormones) and M (musculoskeletal system) were used by between 1 and 10% of the children, and cardiovascular (C), genitourinary (G), anti-neoplastic (L) and anti-parasitic (P) drugs by less than 1%. Among the children aged 2-11 years, the prevalence of use of anti-infectives, respiratory drugs and dermatologicals decreased to respectively 30%, 21% and 17% ; that of the drugs belonging to ATC classes A, H, M, N, P and S was between 1 and 10% ; and that of drugs belonging to class B, C, G and L was less than 1%. In adolescents (12-18 years of age), anti-infectives, respiratory and dermatologicals were used by more than 10% per year. Most of the other drug classes were used by 1-10%, but the prevalence of use of cardiovascular and anti-neoplastic drugs was less than 1%.

EEliflif FatmaFatma SSenen BW.inddBW.indd 2299 003-01-113-01-11 15:1715:17 Table 1: Characteristics of the study population Patient characteristics Children N PY N (% ) Rx N Rx per PY Italy <2 years# 56000 87408 (22) 286597 3.3 2-11 years# 103195 296148 (73) 690688 2.3 12-18 years#* 18154 22599 (6) 35883 1.6

Females 61962 194744 (48) 462580 2.4 Males 67525 211412 (52) 550588 2.6

2000 11188 369 (0) 1150 3.1 2001 73364 45330 (11) 140764 3.1 2002 95712 78850 (19) 220207 2.8 2003 103987 94131 (23) 242261 2.6 2004 106555 96388 (24) 206535 2.1 2005 102911 91086 (22) 202251 2.2 Total 129,487 406,156 (100) 1,013,168 2.5 UK <2 years# 95060 106250 (6) 494353 4.7 2-11 years# 262306 855678 (52) 2011153 2.4 12-18 years#* 229959 683900 (42) 1549372 2.3

Females 219669 804646 (49) 2047616 2.5 Males 225153 841182 (51) 2007262 2.4

2000 307884 288450 (18) 659067 2.3 2001 306923 286483 (17) 677373 2.4 2002 305088 285664 (17) 670690 2.3 2003 303594 280085 (17) 679216 2.4 2004 287287 259219 (16) 674389 2.6 2005 265273 245927 (15) 694143 2.8 Total 444,822 1,645,828 (100) 4,054,878 2.5 Netherlands <2 years# 25694 36601 (13) 78983 2.2 2-11 years# 62326 159010 (56) 208134 1.3 12-18 years#* 40364 87078 (31) 147250 1.7

Females 49709 138262 (49) 230466 1.7 Males 51850 144427 (51) 203901 1.4

2000 56423 48752 (17) 76319 1.6 2001 53274 46822 (17) 76059 1.6 2002 57998 50219 (18) 81919 1.6 2003 62216 49279 (17) 73462 1.5 2004 60315 50882 (18) 75399 1.5 2005 52252 36735 (13) 51209 1.4 Total 101,559 282,689 (100) 434,367 1.5 PY=person-years; Rx=prescriptions;. #The number of children in the various age groups does not add up to the total since one child can contribute to more than one age category. * for Italy the age range included only 12-14 years.

EEliflif FatmaFatma SSenen BW.inddBW.indd 3300 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries- 18 12- 11 < 2 2- 18 12- 11 < 2 2- 18 12- 11 anti-parasitic respiratory sensory < 2 2- 18 12- 11 system nervous < 2 2- 18 12- 11 skeletal musculo- < 2 2- 18 12- 11 anti- neoplastic < 2 2- 18 12- 11 < 2 2- 18 12- ears), and anatomic drug class (X-axis) ears), and anatomic 11 < 2 2- 18 12- 11 < 2 2- genito-urinary hormones anti-infectives 18 12- 11 dermatological < 2 2- 18 12- 11 vascular < 2 2- 18 12- 11 < 2 2- 18 12- 11 alimentary blood cardio- < 2 2- 0

500 450 400 350 300 250 200 150 100 prevalence per 1000 PY 1000 per prevalence One-year prevalence of drug prescriptions by age (< 2, 2-11 12-18 y by 1: One-year of drug prescriptions prevalence Figure

EEliflif FatmaFatma SSenen BW.inddBW.indd 3131 003-01-113-01-11 15:1715:17 Regarding gender diff erences, in the youngest age groups, most of the drugs were equally prescribed to both genders or more frequently prescribed to boys than girls (rate ratio (RR) <1), particularly anti-infectives and respiratory drugs. This pattern reversed in adolescence, when user prevalence for almost all drug classes (except non-sex hormones) was higher among girls than boys. This gender pattern, which was consistent across countries, was most pronounced for genitourinary drugs (G), whose user prevalence was more than 60-fold higher in girls because they include oral contraceptives, which accounted for 95% of the girls’ use of G drugs. The use of drugs for blood and blood forming organs (mainly iron preparations) was also mark- edly higher among adolescent girls. The age trend of prevalence of use was consistent across countries, although there were some variations in the age-specifi c rates (Figure 2). In particular, the UK showed the highest prevalence of alimentary drug use (class A) in children aged less than two years, and the prevalence of prescriptions of dermatological drugs (class D) was 3-4 fold higher in the UK and the NL than in Italy (both p-values <0.001). The prevalence of genitourinary drugs (almost all oral contraceptives) was very high in adolescent girls in the NL (p<0.001). In Italy, the use of hormones (almost all systemic corticosteroids) was 10-fold higher in children aged <2 years (p<0.001) and 5-fold higher in those aged 2-11 years (p<0.001); respiratory drug use was also greater in Italy than in the other two countries (p<0.001). The prevalence of the use of anti-infectives and drugs for musculoskeletal disorders was much lower in the NL; the prescription prevalence of nervous system drugs (including paracetamol that is on prescription in UK) was much higher in the UK; and the use of drugs for the sensory organs was much less in Italy.

1400

1300 UK NL IT 1200 1100 1000 900 800 700 600 500

prevalence of use per 1000 400 300 200 100 0 <2 2-11 12- <2 2-11 12- <2 2-11 12- <2 2-11 12- <2 2-11 12- <2 2-11 12- <2 2-11 12- <2 2-11 12- <2 2-11 12- 18 18 18 18 18 18 18 18 18

Alimentary dermatological genitourinary hormones anti-infectives musculoskeletal nervous system respiratory sensory organs

Figure 2: Year prevalence of drug use (per 1000 PY) by age (<2, 2-11, 12-18 years), database (IMS=UK, IPCI= Netherlands, Pedianet=Italy) and anatomic level for the most prevalently used drug classes (PEDIANET data are excluded for the age category 12-18 years)

EEliflif FatmaFatma SSenen BW.inddBW.indd 3232 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries-

Prevalence of drug use in therapeutic drug class Within the most frequently used anatomic drug classes, the therapeutic class of anti-bacterials (J01) accounted for most of the anti-infective drug use; and the therapeutic classes anti- asthmatics (R03), other respiratory products (R07) and nasal preparations (R01) were the most frequently used drugs in the respiratory group (Table 2). Topical corticosteroids (D07) and emollients and protectives (D02) were the therapeutic classes with the highest prevalence of use among the dermatological drugs. Many therapeutic classes in the group of alimentary drugs (laxatives, anti-diarrheals, drugs for acid disorders) had a considerable prevalence of use. The most prescribed drugs in the other classes were anti-anemia medications (ATC class B03), cardiac therapy (ATC class C01, mainly digoxin), sex hormones (ATC class G03), oral corticosteroids (ATC class H02), non-steroidal antiinfl ammatory drugs (ATC class M01), analgesics (ATC class N02) and ophthalmologicals (ATC class S01). Ranking of age-specifi c user prevalence rates of therapeutic classes over the entire range of drugs showed that antibacterials (J01) are the most frequently prescribed drugs in all age groups (table 3), and are at least prescribed to twice as much children as the number two most frequently used drug in each age category. The second most used drug changed by age from ophtalmologicals (S01) (<2 years), drugs for obstructive airway disease (R03) (2-11 years) to sex hormones (G03)(12-18 years). When the therapeutic classes within each anatomic class were ranked on the basis of the ratio between recurrent (chronic) and non-recurrent(acute), a diff erent pattern was observed (table 2). The drugs with a ratio of >1 (indicating mostly chronic/recurrent use) were often those with a low prevalence of use (except for sex hormones): antidiabetics (A10), digestives (A09), bile and liver therapy (A05), anti-thrombotic agents (B01), agents acting on the renin-angiotensin system (C09), lipid-lowering drugs (C10), sex hormones (G03), thyroid therapeutic agents (H03), immunosuppressive agents (L04), muscle relaxants (M03), anti-epileptics (N03), and psychoanaleptics (N06) (Table 2). In absolute terms, emollients (D02), topical corticosteroids (D07), sex hormones (G03), anti-infectives (J01) and drugs for obstructive airways disease (R03) showed the highest prevalence of recurrent use.

Most frequently used drugs in each anatomic drug class In the most frequently used anatomic classes (D, J and R), the most frequently used individual dermatological drugs were fusidic acid (except for Italy), topical steroids and topical imidazole/ triazole derivatives (table 4). The topical triazoles/imidazoles were off -label in most countries for at least one or more age categories. In the anti-infectives group (J), penicillin derivatives (amoxicillin, amoxicillin+clavulanic acid and phenoxymethylpenicillin) followed by macrolides (erythromycin, clarithromycin) were the most frequently used drugs, cefalexin (UK, < 2 year) was the only off -label drug. Oral acyclovir was one of the top 5 anti-infective drugs in Italy. In the respiratory drug group salbutamol, and inhaled steroids (beclomethasone, fl uticasone, fl unisolide), antihistaminics (cetirizine, loratidine, clorpheniramine) and xylometazoline were

EEliflif FatmaFatma SSenen BW.inddBW.indd 3333 003-01-113-01-11 15:1715:17 Table 2: Prevalence of acute use (<3 prescriptions per year) and recurrent use (>3 prescriptions per year) by age and therapeutic level (prevalence per 1000 PY), ranked by the ratio of recurrent/acute use prevalence.4 Acute use prevalence per Recurrent use Ratio re- Total 1000 PY prevalence per 1000 PY current/ preva- acute lence ATC Explanation <2 2-11 12-18 All <2 2-11 12- All yrs. yrs. yrs. ages yrs. yrs. 18 ages yrs. A Alimentary tract A10 Drugs used in diabetes 0.0 0.2 0.3 0.2 0.0 0.9 2.5 1.3 7.0 1.5 A09 Digestives, including 0.1 0.0 0.0 0.0 0.1 0.2 0.2 0.2 4.9 0.2 enzymes A05 Bile and liver therapy 0.1 0.0 0.0 0.0 0.1 0.1 0.1 0.1 1.7 0.1 A12 Mineral supplements 1.1 0.8 0.5 0.7 0.2 0.3 0.2 0.2 0.3 1.0 A06 Laxatives 24.7 13.3 6.2 12.0 3.3 4.7 1.8 3.6 0.3 15.6 A02 Drugs for acid-related 27.0 3.5 9.6 7.9 12.6 0.8 1.7 2.3 0.3 10.1 disorders A11 Vitamins 24.2 3.6 1.4 4.9 3.5 0.7 0.5 0.9 0.2 5.8 A04 Anti-emetics and anti- 1.4 0.6 3.7 1.8 0.7 0.1 0.2 0.2 0.1 2.0 nauseants A03 Drugs for functional 25.9 10.6 9.7 11.8 1.4 0.4 0.9 0.6 0.1 12.4 gastrointestinal disorders A01 Stomatological 56.3 6.6 4.2 10.7 3.2 0.2 0.2 0.5 0.0 11.2 preparations A07 Antidiarrheals 64.9 11.6 3.2 14.0 1.9 0.3 0.6 0.5 0.0 14.5 B Blood and blood forming organs B01 Anti-thrombotic 0.2 0.2 0.3 0.3 0.2 0.3 0.3 0.3 1.1 0.5 agents B03 Anti-anemic 20.8 3.6 6.8 6.4 2.9 0.5 1.2 1.0 0.2 7.4 preparations B02 Anti-hemorrhagics 5.5 1.0 1.6 1.6 0.2 0.1 0.2 0.1 0.1 1.8 C Cardiovascular system C09 Agents acting on the 0.1 0.1 0.1 0.1 0.1 0.2 0.3 0.2 2.5 0.3 rennin-angiotensin system C10 Lipid modifying 0.0 0.0 0.1 0.0 0.0 0.1 0.1 0.1 1.7 0.1 agents C03 Diuretics 0.6 0.1 0.1 0.2 0.6 0.1 0.2 0.2 1.2 0.4 C08 Calcium channel 0.0 0.0 0.2 0.1 0.0 0.0 0.2 0.1 0.8 0.2 blockers C07 Beta blocking agents 0.1 0.2 2.2 0.8 0.1 0.2 0.7 0.3 0.4 1.2 C01 Cardiac therapy 1.3 2.2 1.6 1.9 0.2 0.3 0.3 0.3 0.2 2.2 D Dermatologicals D10 Anti-acne preparations 0.3 1.0 31.2 11.2 0.0 0.1 15.0 5.2 0.5 16.3 D02 Emollients and 98.8 45.1 25.6 43.8 48.5 21.8 8.1 19.8 0.5 63.6 protectives D05 Anti-psoriatics 3.9 3.1 4.8 3.7 0.2 0.5 2.0 1.0 0.3 4.7 D07 Corticosteroids, 140.4 74.1 55.9 74.4 24.4 11.9 8.7 12.0 0.2 86.5 dermatological preparations

EEliflif FatmaFatma SSenen BW.inddBW.indd 3344 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries-

Table 2: Continue

Acute use prevalence per Recurrent use Ratio re- Total 1000 PY prevalence per 1000 PY current/ preva- acute lence ATC Explanation <2 2-11 12-18 All <2 2-11 12- All yrs. yrs. yrs. ages yrs. yrs. 18 ages yrs. D03 Preparations for 1.1 0.7 1.0 0.8 0.0 0.1 0.1 0.1 0.1 0.9 treatment of wounds and ulcers D08 Antiseptics and 3.9 2.4 2.8 2.7 0.1 0.1 0.2 0.1 0.1 2.8 disinfectants D01 Anti-fungals for 50.8 18.4 19.6 22.0 1.6 0.6 1.5 1.0 0.0 23.0 dermatological use D06 Antibiotics and 43.6 36.4 23.6 32.8 0.8 0.9 0.9 0.9 0.0 33.7 chemotherapeutics D11 Other dermatological 5.3 8.9 9.8 8.9 0.2 0.1 0.4 0.2 0.0 9.1 preparations G Genito-urinary system and sex hormones G03 Sex hormones, 1.7 0.4 32.3 11.3 0.3 0.1 49.7 17.0 1.5 28.3 modulators of the genital system G04 Urologicals 0.5 1.1 1.8 1.3 0.1 0.6 0.6 0.5 0.4 1.8 G01 Gynecological 1.1 1.3 9.2 4.0 0.0 0.0 0.5 0.2 0.0 4.2 anti-infectives and antiseptics H Systemic hormonal preparations, excluding sex hormones and insulins H03 Thyroid therapy 0.3 0.2 0.3 0.2 0.4 0.5 1.1 0.7 3.1 0.9 H01 Pituitary and 0.1 2.2 1.5 1.7 0.0 1.3 1.4 1.2 0.7 3.0 hypothalamic hormones H02 Corticosteroids for 51.0 23.2 8.0 20.7 6.0 2.2 1.0 2.2 0.1 22.9 systemic use H04 Pancreatic hormones 0.0 0.3 0.7 0.4 0.0 0.0 0.1 0.0 0.1 0.4 J Anti-infectives for systemic use J01 Anti-bacterials for 340.0 241.4 166.3 225.6 95.2 47.0 27.6 45.2 0.2 270.7 systemic use J04 Anti-mycobacterials 0.5 0.5 0.3 0.4 0.1 0.0 0.0 0.0 0.1 0.5 J07 Vaccines (exc. 11.8 10.6 14.3 12.0 0.8 0.4 1.0 0.6 0.1 12.6 routine childhood vaccinations) J02 Antimycotics for 1.1 0.6 3.7 1.7 0.0 0.0 0.2 0.1 0.0 1.8 systemic use J05 Antivirals for systemic 9.8 4.2 1.7 3.9 0.1 0.0 0.1 0.1 0.0 4.0 use

EEliflif FatmaFatma SSenen BW.inddBW.indd 3355 003-01-113-01-11 15:1715:17 Table 2: Continue

Acute use prevalence per Recurrent use Ratio re- Total 1000 PY prevalence per 1000 PY current/ preva- acute lence ATC Explanation <2 2-11 12-18 All <2 2-11 12- All yrs. yrs. yrs. ages yrs. yrs. 18 ages yrs. L Anti-neoplastic and immunomodulating drugs L04 Immunosuppressive 0.0 0.0 0.1 0.1 0.0 0.1 0.4 0.2 3.8 0.3 agents L01 Anti-neoplastic agents 0.0 0.1 0.1 0.1 0.0 0.1 0.1 0.1 0.9 0.2 M Musculoskeletal system M03 Muscle relaxants 0.1 0.1 0.2 0.1 0.0 0.1 0.2 0.2 1.7 0.2 M01 Anti-infl ammatory 38.8 32.0 53.6 40.0 1.2 1.2 3.0 1.8 0.0 41.8 and anti-rheumatic products N Nervous system N03 Anti-epileptics 0.7 0.7 0.8 0.7 1.1 2.6 3.6 2.8 3.9 3.5 N06 Psychoanaleptics 0.1 1.1 6.8 2.9 0.0 1.7 6.8 3.3 1.1 6.2 N04 Anti-parkinson drugs 0.0 0.0 0.1 0.0 0.0 0.0 0.1 0.0 1.1 0.1 N05 Psycholeptics 7.3 2.2 5.0 3.6 0.4 0.4 1.6 0.8 0.2 4.5 N07 Other nervous system 0.1 0.2 2.0 0.8 0.0 0.0 0.4 0.2 0.2 1.0 drugs N02 Analgesics 109.9 55.0 38.7 54.9 24.2 8.5 5.6 9.0 0.2 63.9 N01 Anaesthetics 2.1 4.2 4.2 4.0 0.1 0.1 0.2 0.1 0.0 4.1 P Antiparasitic products P03 Ectoparasiticides 2.9 14.9 10.6 12.2 0.1 1.5 0.8 1.1 0.1 13.4 P01 Anti-protozoals 1.8 1.7 2.2 1.9 0.0 0.0 0.0 0.0 0.0 1.9 P02 Ant-helmintics 4.4 12.2 3.1 8.4 0.1 0.2 0.0 0.1 0.0 8.5 R Respiratory system R03 Drugs for obstructive 126.3 69.3 39.2 64.7 34.8 39.1 31.8 36.2 0.6 100.9 airway diseases R07 Other respiratory 45.4 55.1 53.0 53.4 2.8 8.1 14.1 9.6 0.2 63.0 system products R06 Anti-histamines for 50.4 29.1 17.4 27.3 3.5 2.1 2.6 2.4 0.1 29.7 systemic use R01 Nasal preparations 79.1 36.2 43.7 43.0 3.9 2.1 4.4 3.1 0.1 46.1 R05 Cough and cold 4.1 2.2 1.7 2.2 0.2 0.1 0.0 0.1 0.0 2.3 preparations R02 Throat preparations 1.1 1.4 4.0 2.3 0.0 0.0 0.1 0.0 0.0 2.3 S Sensory system S01 Ophthalmologicals 164.9 60.7 42.9 64.9 10.3 3.1 4.0 4.1 0.1 69.0 S03 Ophthalmological 3.2 3.5 4.0 3.7 0.1 0.2 0.1 0.1 0.0 3.8 and otological preparations S02 Otologicals 15.7 15.0 13.5 14.6 0.4 0.5 0.6 0.5 0.0 15.1 All ATC therapeutic levels with prevalence of both acute and recurrent use below 0.1 per 1000 PY have been excluded.

EEliflif FatmaFatma SSenen BW.inddBW.indd 3366 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries-

Table 3: Top ten of most frequently used therapeutic classes in various age categories < 2 years Therapeutic class users per 1000 PY J01 Antibacterials for systemic use 435 S01 Ophthalmologicals 175 D07 Corticosteroids, dermatological preparations 165 R03 Drugs for obstructive airway diseases 161 D02 Emollients and protectives 147 N02 Analgesics 134 R01 Nasal preparations 83 A07 Antidiarrheals, intestinal anti-infl ammatory/anti-infective agents 67 A01 Stomatological preparations 59 H02 Corticosteroids for systemic use 57 2-11 years Therapeutic class users per 1000 PY J01 Antibacterials for systemic use 288 R03 Drugs for obstructive airway diseases 108 D07 Corticosteroids, dermatological preparations 86 D02 Emollients and protectives 67 S01 Ophthalmologicals 64 N02 Analgesics 63 R07 Other respiratory system products 63 R01 Nasal preparations 38 D06 Antibiotics and chemotherapeutics 37 M01 Anti-infl ammatory and antirheumatic products 33 12-18 years Therapeutic class users per 1000 PY J01 Antibacterials for systemic use 194 G03 Sex hormones and modulators of the genital system 82 R03 Drugs for obstructive airway diseases 71 R07 Other respiratory system products 67 D07 Corticosteroids, dermatological preparations 65 M01 Anti-infl ammatory and antirheumatic products 57 R01 Nasal preparations 48 S01 Ophthalmologicals 47 D10 Anti-acne preparations 46 N02 Analgesics 44

most frequently prescribed. Beclomethasone, xylometazoline and cetirizine were off -label in the youngest children (< 2 years) in either UK or NL. In the moderately used drugs (anatomic groups A, G, N, S), the most frequently prescribed alimentary tract drugs (A) were laxatives (lactulose), nystatin, domperidone, mebeverine or ranitidine. Only ranitidine and laurilsulfate were off -label in children < 2 years of age. For the genito-urinary drugs, oral contraceptives and topical anti-fungals (miconazole) determined the top fi ve in NL and UK, whereas Italy (only until age 12) had estrogens, drugs for treatment of incontinence and antiseptics as most frequently pescribed drugs. The percentage of off -label

EEliflif FatmaFatma SSenen BW.inddBW.indd 3377 003-01-113-01-11 15:1715:17 total total users # OL years # % OL 341 0 124 0 465 try uoride 571 0 192 0 763 Drug name < 2 years 2-11 bromide users ## % % ## OL >12 years Total # % OL # % OL # % 12 100 52 100 37 0 103 Phytomenadione 88 0 9 0 97

Drug name < 2 years 2-11 years acid WarfarinPhytomenadione 26 0 10 1 0 7 100 17 0 43 100 25sulfate 100 46 Ferrous 0 gluc. Na Ferrous 48 6 0 48 0 62 0 68 users # OL >12 years Total # % OL # % OL # % 12 100 29 100 10 100 51 Epinephrine 6 100 580 0 383 0 970 Epinephrine 16 0 56 0 72 1 100 12 100 0 Na 13 Acetylsalicylic NLDrug Name < 2 years 2-11 years UK IT Cardiovascular (C) Cardiovascular Hydrocortisone (hemorrhoids) LidocainePropranololEpinephrine 3Enalapril 030 0 100 Na 0 13 5 Na 0 46 0 17 0 18 Propranolol Na 0 4 0 23 2 4 Furosemide 0 0 5 21 0 27 7 0 18 Atenolol38 0 262 100 Enalapril 0 293 0 19 Hydrocortisone 0 76 2 0 0 Furosemide 100 31 Na 14 Na 100 42 9 0 14 26 100 78 0 4 37 0 Oxetacaine 0 0 63 13 Disopyramide 0 Na 0 8 Na 0 1 8 0 1 Alimentarry tract (A) LactuloseDomperidoneMiconazole 79 92 0 222 0 332Nystatin 200 0 73 0 58Laurilsulfate 0 30 0 374fumarate 0 482Blood and blood forming organs (B) 0 8 2 Miconazole Lactulose 20Ferrous 13080 0 11 0 238 0 60Phytomenadione 100 41 0 17 566 Ranitidine Carbasalate 0 3 0 57 0 2 797 0 134 0 117calcium acid 0 119 0 2565 0 144 0 31 Domperidone 0 565 0 3 145 Folic 133 Mebeverine 0 731 0 3927fl 103 100 0 343 0 46 Domperidoneacid0 141 0 136 Sodium 48 0 622 0 100 Tranexamic 0 247 250 0 368 Cimetropium 0 649 0 486 Na Na 0 558 0 899 57 Lactitol 9 Electrolytes 0 524 0 295 0 581 0 304 124acid4 0 151 Nystatin 45 0 Tranexamic 0 168 0 213 0 168 275 0 139 172 0 133 0 272 Table 4: Most frequently used drugs per anatomic level by country and age in 2005 plus paediatric licensing status in each coun status licensing country and age in 2005 plus paediatric by level used drugs per anatomic 4: Most frequently Table

EEliflif FatmaFatma SSenen BW.inddBW.indd 3388 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries- 205 0 431 0 636 57 0 26 0 83 antibiotics estrogens 0 Na 1 100 693 0 695 Povidone-iodine 1 100 14 100 15 estrogen 1 100 30 100 1034 Na 100 1038 4 Clotrimazole 100 321 61 100 325 100 182 Norethisterone 100 801 0 0 1046 Na Conjugated 4 100 1019 100 1025 Oxybutynin 0 Na 37 0 37 Lynestrenol 0 Na 4 100 57 100 61 Desogestrel 1 100 0 100 268 100 272 Estriol 9 100 5 100 14 Dermatological (D) Dermatological acid Fusidic 194Hydrocortisone 100 1013Miconazoleacid 284734 100 3112691287Fusidic 100 100 Triamcinolone 100 1518 100 HydrocortisoneKetoconazole 880 273 36 2425 0 337Genito urinary system and sex hormones (G) 0 3936 0 100 360 48 0 1457 0 204 7311 / Levonorgestrel 0 6273 100 292 0 100 168 0 814estrogen Mometasone 100 688 2574 100 139 Clobetasone 0 / Cyproterone Clotrimazole 100 355 12310/ estrogen 240 Betamethasone 0 362 Betamethasone 232Norethisterone 0 602 828 74 0 1888Miconazole 100 1617 0 0 1080 0 0 3200 0 627 967 Na Mupirocin 4 0 2 0 3073 1360 100 14 0 Clotrimazole 100 98 90 2401 100 58 0 313 100 Econazole 0 403 175 100 76 Levonorgestrel 100 118 100 293 prog/ 0 Medroxy 90 100 83 Na 100 173 0 0 946 0 946 Benzydamine 2 100 19 100 21

EEliflif FatmaFatma SSenen BW.inddBW.indd 3399 003-01-113-01-11 15:1715:17 users # OL # % OL 0 6 0 24 5 0 17 0 22 1760 0 4210 0 5970 umate 118 100 446 100 564 xamethasone 18 Drug name < 2 years 2-11 years total sodium /clavulanic users ## % % ## OL >12 years Total # % OL # % 44 0 8 0 72 De OL # % 518 0 6057 0 5710 0 12285 Amoxicillin 2573 0 3603 0 6176 Drug name < 2 years 2-11 years

penicillin Total Total users # OL 12 > years # % OL # % OL # % 1 0 13 0 16 0 30 Glucagon 0 Na 77 108 0 0 185 Desmopressin 0 Na 120 0 120 133 0 657 0 155 0 945 Flucloxacillin 897 0 6043 0 4223 0 11163 Amoxicillin cont’d NLDrug name < 2 years 2-11 years UK IT Systemic hormonal preparations (H) hormonal preparations Systemic DesmopressinPrednisolone 0 Levothyroxine Na 14sodium 94 100 41Prednisone 49 0 100 31 0 143 100 86 0 Desmopressin Levothyroxine Na 0 11 19 100 7 Na 0 467 89 100 18 0 312 0 Dexamethasone 0 779 159 19 0 Betamethasone 100 267 1430 0 2064 0 3494 Prednisone 5 0 240 0 245 Amoxicillin / Amoxicillin clavulanic Dexamethasone 4Antiinfectives for systemic use (J)\ 0 6Amoxicillin 0 2 0 12 763 0 1870 Somatropin 0 302Clarithromycin 0 2935 Phenoxymethyl Azithromycin 131 0 0 489Pheneticillin 47 0 137 Na agents (L) Antineoplastic and immunomodulating 0 757 0 246 28 22Fluorouracil Erythromycin 0 111 0 211 0 404Azathioprine 0 26 1287 0 161 Trimethoprim 0Triptorelin 0 394 0 5265 0 54 0Methotrexate 0 3386 Na Cefalexin 351 0 9938 Na 6Ciclosporin 0 2623 Levothyroxine 0 0 Azithromycin 0 system (M) 0 2122Musculo-skeletal 100 3 Na 0 5096 Na 345 666Diclofenac1597 0 0 3 2 Clarithromycin 1 0 2616 100 100 9 0 1098Naproxen 0 3282 Na 683 0 3 0 3041 100 0 0 0 1 0 2385 0 Aciclovir Azathioprine 0 3068 100 2 Methotrexate 0 1 0 Na 0 0 29 0 Na Ciclosporin 1 0 Tacrolimus 309 10 0 233 Na 0 739 0 262 Na 16 Goserelin 0 1048 171 0 0 10 Ibuprofen 0 0 181 0 Diclofenac Na 0 24 65 Na 0 13 5 1085 0 34 0 0 5404 2 81 Na 100 8 0 4251 Leuprorelin 0 9 0 210740 Ibuprofen 41 0 Pidotimod 0 2 0 14 0 22 100 0 12476 11 0 1290 5081907 100 1399100 Methotrexate na Triptorelin 100 Mornifl 0 7 Ciclosporin 0 807 0 100 0 na 0 91 5 na na5 6 100 3 100 6 3 100 Table 4 Table

EEliflif FatmaFatma SSenen BW.inddBW.indd 4400 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries- 255 0 506 0 761 umic acidumic 62 0 168 0 232 combinations combinations 0 404 Diazepam 41 0 85 0 126 398 286 0 433 0 719 Paracetamol, ylphenidate 0 Na 0 169 Meth 125 0 140 0 265 Paracetamol 4292 0 11085 0 2832 0 18209 Paracetamol 603 0 491 0 1094 0 Na 2 100 12 100 14 Naproxen 0 Na 4 0 143 0 147 Flurbiprofen 19 0 220 0 239 3 0 1100 0 14 0 127 Na 27 Pizotifen 0 79 0 106 0 Fluoxetine Na 207 0 0 430 Na 0 637 6 Niaprazine 0 158 0 39 0 197 Ibuprofen Diclofenac, combinations 0Bufexamac NaNervous system (N) 29Methylphenidate 3 0 0 131Paracetamolacid0 0 160 100 8 Na Mefenamic 38 100 3 0 99 Na 100 14 1132 0 Ketoprofen 0 1278 0 1289 Ketoprofen 0 Na 8 15 0 354 100 70 0 362 0 86 Nifl Diazepam 83934 100 100 0 81 Diazepam 4 0 124 0 266 0 394acid Valproic 4 0 39 0 43 Lidocaine/ prilocaine Carbasalate calcium

EEliflif FatmaFatma SSenen BW.inddBW.indd 4411 003-01-113-01-11 15:1715:17 # users OL # % OL 42 100 229 100 271 90 0 75 0 165 537 0 7251262 0 117 0 256 0 373 erent erent oquine 8 0 16 0 24 Permethrin 0 na 13 13 0 combinations combinations and antiinfectives of diff antibiotics Drug name < 2 years 2-11 years total ## % % users ## OL >12 years Total # % OL # % OL # %

Hydrocortisone 101 100 236 0 57 0 395 Combinations Drug name < 2 years 2-11 years Total Total users # OL # % >12 years OL # % age -label for OL 0 Na 4 0 10 0 14 Permethrin 35 0 845 0 400 0 1280 Mefl 12 0 12970 0 acid 0 211 Fusidic 13161951 0 0 5403807 0 Dexamethasone # % cont’d Antiparasitic drugs, insectides and repellents (P) drugs, Antiparasitic MebendazoleMetronidazole Proguanil, 1 2combinations 0 100 21 87Permethrin20 0 14 0 0 43Respiratory system (R) 0 Phenothrin 102 1 MebendazoleSalbutamol 0 8Fluticasone3 311Desloratadine 24 0 31695 100 12 0 159Xylometazoline 0 1053 Malathion 14 0 349201 0 154Levocetirizine 0 702448 100 356 0 2069 0 0 53 0 447201 Mebendazole143 0 0 0 366257 1813 0 Sensory organs (S) 0 0 1062acid Pyrantel 0 Salbutamol 0 Beclometasone 40 654 Na 0 Fusidic 827 Chlorphenamine1088 38372 Cetirizine 1771500Albendazole 0 Levocabastine 256 342 0 578 0 479 1309 0 100 4416332 100 263 0 30212403 100 1049 100 3963 100 3945 0 Chloramphenicol 0 11 0 2Hydrocortisone 517 10552 0 8321 0 4155 100 130 0 156 0 959 Salbutamol 479/antiinfectives 145 100 7161 100 acid 1 100 24 0 22034 0 291 Loratadine5482 0 532192 Beclometasone3382 Cromoglicic 100 156 0 22 Cetirizine 0 1875 100 13509 0 4145 1202 1584 0 Lidocaine 0 Tobramycin263023 0 0 7552 0 1932 0 4559 0 2849 0 Flunisolide Antiinfectives, 3134 1 0 234 441 0 4433 331435 0 100 135 0 0 515 615 19921669 0 16 0 1256 0 2261 0 956 0 1871 0 0 0 185 4254 Nedocromil Salbutamol 0 Na 265 0 465 0 730 Nedocromil 43 0 156 0 199 NLDrug name < 2 years 2-11 years UK IT Table 4 Table Na=not assessable ; OL= off

EEliflif FatmaFatma SSenen BW.inddBW.indd 4422 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries-

use of oral contraceptives and antifungals was high in NL and UK. In the group of nervous system drugs paracetamol is clearly the most used (but underestimated due to high OTC use in NL and IT) drug, methylphenidate (NL and UK), lidocaine (NL), pizotifen (UK), fl uoxetine (UK) diazepam, niaprazine (IT) and valproic acid (IT) were also in the top 5 of at least one country. None of them was off label, except diazepam < 12 years of age in NL. In the group of sensory organ drugs many diff erent drugs were used in the various countries, the most frequently prescribed drugs in NL (fusidic acid, levocabastine) and UK (chloramphenicol) were off label. The low prevalence drugs comprised many classes (groups B, C, H, L. M, P). In the blood forming organs group (B), phytomenadione, iron, tranexamic acid, platelet inhibitors and vitamin K antagonists were most frequently prescribed. Salicylic acid derivatives were off -label. In the cardiovascular drug group topical steroids (anti-hemorrhoidal crèmes), topical anesthetics (lidocaine, oxetacaine), B-blockers (propanolol, atenolol), furosemide, disopyramide, epinephrine and enalapril were most frequently prescribed. Furosemide, B-blockers, epinephrine and topical (antihemorrhoidal) steroids were off -label in at least one country. For the non-sex hormones (H) desmopressin, oral steroids, (dexamethasone, prednisolone and prednisone), levothyroxine and glucagons) were the most frequently prescribed drugs. Only the oral steroids were off -label (NL and UK only). The most frequently prescribed anti-neoplastic and immunomodulating drugs (L) diff ered substantially between countries but were almost always off -label. In the musculoskeletal drug group (M) NSAIDs were the most frequently prescribed drugs, with important sequence diff erences between countries but little off -label use except in Italy where the number one and two drugs (ibuprofen and mornifl umate) were off -label. The number one anti-protozoal drug (P) was mebendazole in all countries, there was very little off -label drug use.

DISCUSSION

This study describes age-specifi c paediatric drug prescribing in three diff erent European countries from a very high aggregation level of anatomic classes to the level of individual drugs. This information could facilitate the prioritisation of paediatric research by regulatory authorities.

Prioritisation of drug safety research in paediatrics Prioritising the research needs in medicines for children has currently been done mostly on the basis of one component, whereas we recommend two important assessments: The fi rst is public health assessment16 comprising the severity and prevalence of disease and the availability of treatment alternatives. In addition to the public health assessment a second assessment should be done which is utilisation assessment. Utilisation assessment may comprise the frequency or volume of use and the licensing/labelling status of medicines for children. The use of off -label and unlicensed medicines implies that there are no proper labelling and dosing

EEliflif FatmaFatma SSenen BW.inddBW.indd 4433 003-01-113-01-11 15:1715:17 recommendations, which can potentially be harmful to children 17 18 19 20. Therefore off -label and unlicensed medicines should be of higher priority for research than licensed/on-label medications, especially if no data on safety, effi cacy in children is available. Our study focused on assessing the volume and labelling status in order to provide knowledge to experts and facilitate research prioritisation that includes both the public health as well as the utilisation assessment. The utilisation data from this study underline the conclusions of the recently published EMEA consensus/expert derived list of research priorities concerning off -patent medicinal products16, which emphasized the need for paediatric studies of the safety of topical, systemic and inhaled steroids. Steroids are associated with impairment of growth21, abnormalities in the metabolism of glucose22 and adrenal suppression23 24. Of these adverse eff ects, growth retardation is the most common and is of particular concern in children. The extent of growth suppression varies with the method of administration (e.g., inhaled or oral) and the duration of treatment as well as with the type and dose of glucocorticoid used21 25. Topical and systemic antifungals (imidazoles/triazoles), acid-reducing drugs and antineoplastic drugs are also listed as research priorities by EMEA. Our utilization data found that these drugs are frequently or recurrently used and are mostly off -label. Many other drugs that are listed on the priority list of EMEA did not appear as frequently used drugs in this study, and would not be considered priorities on the basis of frequency of use in primary care alone, but apparently were considered priorities for other reasons. On the other hand sex hormones are not listed on the priority list of the EMEA, whereas they are frequently and recurrently prescribed, and mostly off -label. Few long term safety studies on the use of sex hormones in adolescents are available and to our knowledge no RCTs on the safety and effi cacy of sex hormones in adolescents have been conducted. The use of oral anticonceptives in adolescents has been associated with an increased risk of lower bone mineral density, higher serum cholesterol and triglyceridemia26-28. Similar to adults, the use of oral anticonceptives by teenagers has also been associated with an increased risk of cardiovascular events (eg myocardial infarction, stroke) as well as an increased risk of venous thrombo-embolism29-33. As the use of sex hormones in young adolescents is relatively high, implying a long duration of use, further studies on the effi cacy and long term safety eff ects of these drugs when used by youngsters are warranted. Although drug use patterns and labeling status can inform decisions on prioritization of research, these data inform us also about suboptimal use and even may demonstrate undesirable prescribing practices. For example fusidic acid and chloramphenicol are frequently used and often off -label (table 4). Fusidic acid is known to be prescribed for the treatment of conjunctivitis in the Netherlands similarly to chloramphenicol which is used in the UK. However, the benefi cial eff ect of antibiotics in the treatment of this condition is not convincingly proven34 35. Indeed the literature shows that acute bacterial conjunctivitis is frequently a self-limiting condition and topical antibiotic use off ers only marginal benefi t in improving clinical outcomes; hence the emphasis should be on educating clinicians not to prescribe such treatment rather than a

EEliflif FatmaFatma SSenen BW.inddBW.indd 4444 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries-

call for more research 36 37. Another example underlining the need for education rather than research is the cough and cold medications. These drugs are not only available over the counter but are also frequently prescribed which should be strongly discouraged due to reports of death and lack of effi cacy38.

Drug use patterns We show that the prevalence of the most frequently prescribed drugs in primary care is highest in children aged less than two years, that the most frequently used drugs (anti-infectives, dermatolologicals and respiratory drugs) are the same in all three age categories, and that almost all other drugs are used by less than 10% of children per year. In general, three groups can be distinguished: drugs used by more than 10% of children per year, those used by 1-10%, and those by less than 1%. The use of the high prevalence drug classes decreases with increasing age but remains high, whereas the use of the lowest prevalence drug groups increases to a moderate prevalence rate in adolescence except in the case of cardiovascular and anti-neoplastic agents. Only a few therapeutic drug classes accounted for the majority of use in a specifi c anatomic drug class: anti-bacterials, topical corticosteroids, anti-asthma and anti-anemia medications, cardiac drugs, sex hormones, oral corticosteroids, non-steroidal anti-infl ammatory drugs, analgesics and ophthalmologicals. The high prevalence drugs were relatively more often used for acute use. Only 12 drug classes (anti-diabetics, digestives, bile and liver therapy, anti-thrombotic agents, drugs aff ecting the renin-angiotensin system, lipid- lowering drugs, sex hormones, thyroid therapeutic agents, immunosuppressive agents, muscle relaxants, anti-epileptics and psychoanaleptics) were prescribed more frequently for recurrent than acute use. An interesting age-related gender reversal was observed: female adolescents had consistently higher drug use prevalence rates than adolescent boys (except in the case of non-sex hormones), whereas the opposite was true in the younger age categories this was in line with a previous Dutch and Danish study39 40. Interestingly, the percentage of off -label use varied highly between countries and similar drugs diff ered in off -label status between countries. This confi rms that the diff erences in the Paediatric Status of the drugs, instead of the diff erent prescription habits or medical cultures as postulated by many authors represent the real reason for the variability reported by years and from many European studies and surveys on the off -label use in children.41

Previous studies Our study was population-based, had a very large sample size, and covered diff erent European countries. Previous European studies have been country or region specifi c, and have concen- trated on specifi c conditions except for previous Swedish, Dutch and Danish studies in the late nineties and a recent Italian study covering 2000-2006 data40 42-44. These studies all took all types of drugs into account but the methodologies to calculate prevalence and ranking (on

EEliflif FatmaFatma SSenen BW.inddBW.indd 4455 003-01-113-01-11 15:1715:17 the basis of number of dispensed boxes or user prevalence) and age ranges vary largely which complicates direct comparisons. The overall results: highest drug use in lowest age category, ranking of the most frequently used drugs (anti-infectives, respiratory and dermatological drugs), and the gender pattern (more prescriptions for girls than boys after the age of 10) are consistent with the fi ndings of our study39 45 46.

Potential of multi-country database studies Our study shows the potential of studying the primary care prescribing of a wide range of drugs using multiple databases. As all databases include outcome data such as morbidity and mortality data, they can also be used for studies of paediatric drug safety. The country-specifi c estimates provide insights into prescription diff erences and allow a search for high prevalence countries regarding drug prescribing.

Limitations One limitation of this study is the fact that it captured only outpatient, primary care drug prescriptions and no use of over the counter drugs (which has resulted in a substantial underestimation of the use of paracetamol and phytomenadione, and potentially other drugs such as cough and cold medications). At least in the Netherlands, the UK and Italy, most health problems are dealt with in primary care8 and, as drug prescriptions by a specialist for a chronic disease are often continued by general practitioners or paediatricians, most of them are picked up. However, hospital medication administration and the monitoring of chemotherapeutic and biological drugs are unlikely to be fully captured by our databases. Despite diff erences in the absolute prevalence rates of drug prescribing and the types of drugs that were prescribed, age and gender patterns were very consistent in the three countries considered in this study. However, as the UK accounted for 60 percent of the study population, the pooled results are inevitably dominated by UK prescription patterns, therefore stratifi ed analyses were conducted as much as possible. Due to the nature of the databases, we studied drug prescriptions rather than drug intake, and so the prevalence of actual drug exposure may be lower than estimated here. In brief, this paper provides a unique overview of primary care prescription patterns in a large multinational European paediatric population. The data could be used to improve the prioritisation of research into long-term paediatric drug safety, as well as effi cacy and eff ectiveness studies in paediatric medicine. Off -label use in some of the most frequently and recurrently used drugs is high (e.g. oral contraceptives) and these should be considered for prioritisation.

EEliflif FatmaFatma SSenen BW.inddBW.indd 4466 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries-

References

1 Pandolfi ni C, Bonati M. A literature review on off -label drug use in children. Eur J Pediatr 2005;164:552- 558. 2 Regulation (EC) No 1901/2006 of the European Parliament and of the Council of 12 December 2006 on medicinal products for paediatric use. Paediatric regulation. In: Parliament E, editor: Offi cial Jour- nal of the European Union, 2006:18. 3 Sutcliff e A. Testing new pharmaceutical products in children. BMJ 2003;326:64-5. 4 Paediatric exclusivity labelling chnages as of January 5, 2005. http://www.fda.gov/cder/pediatric/ labelchange.htm, 2005. 5 Davies A, Bateman M, Yates A, Bruno M. Pediatric regulations in Europe & US. Regulatory Aff airs focus 2005(May):18-21. 6 Watson R. EU off ers incentives to fi rms to produce medicines for children. BMJ 2006;332:1352. 7 World Health Organization. Make medicines child size, 2007 http://www.who.int/childmedicines/en/ accessed May 4 2008. 8 van der Lei J, Duisterhout J, Westerhof H, van der Does E, Cromme P, Boon W, et al. The introduction of computer-based patient records in the Netherlands. Ann Intern Med 1993;119(10):1036-1041. 9 Sturkenboom M. Other European databases for pharmacoepidemiology. . In: Mann RD AE, editor. Pharmacovigilance`. second ed. London: Wiley, 2007. 10 Sturkenboom M, Nicolosi A, Cantarutti L, Mannino S, Picelli G, Scamarcia A, et al. Incidence of muco- cutaneous reactions in children treated with nilfumic acid, other nonsteroidal antiinfl ammatory drugs, or nonopioid analgesics. Pediatrics 2005;116:e26-33. 11 Vlug A, van der Lei J, Mosseveld B, van Wijk M, van der Linden P, Sturkenboom M, et al. Postmarketing surveillance based on electronic patient records: the IPCI project. Methods Inf Med 1999;38:339-344. 12 ‘t Jong G, Eland I, Sturkenboom M, Anker Jv, Stricker B. Unlicensed and off -label prescription of drugs to children; a population based cohort study. BMJ 2002. 13 Wong I, Murray M. The potential of UK clinical databases in enhancing paediatric medication research. Br J Clin Pharmacol. 2005;59(6):750-755. 14 EMEA. ICH topic 11: Clinical investigation of medicinal products in the pediatric population, 2001. 15 Neubert A, Bonifazi A, Catapano M, Baiardi P, Guiaquinto C, Knibbe C, et al. Defi ning Off -label and Unlicensed Use of Medicines for Children: Results of a Delphi survey. Pharmacol Res 2008;submitted. 16 EMEA. Priority list of off -patent medicinal products for pediatric studies. Doc. ref. EMEA/49677/2006 Corr ed. London: EMEA, 2006. 17 Evidence of harm from off -label or unlicensed medicines in children MEA/126327/2004. European Medicines Agency Pre-authorisation Evaluation of Medicines for Human Use. London October 2004. 18 Horen B, Montastruc J, Lapeyre-Mestre M. Adverse drug reactions and off -label drug use in paediatric outpatients. Br J Clin Pharmacol 2002;54:665-670. 19 Neubert A, Dormann H, Weiss J, Egger T, Criegee-Rieck M, al. RWe. The impact of unlicensed and off -label drug use on adverse drug reactions in paediatric patients. Drug Saf 2004;27:1059-67. 20 Choonara I, Conroy S. Unlicensed and off -label drug use in children: implications for safety. Drug Safety 2002;25:1-5. 21 Allen D. Growth suppression by glucocorticoid therapy. . Endocrinol Metab Clin North Am 1996;25:699- 717. 22 Eigen H, Rosenstein B, FitzSimmons S, Schidlow D. Cystic Fibrosis Foundation Prednisone Trial Group. A multicenter study of alternate-day prednisone therapy in patients with cystic fi brosis. . J Pediatr 1995;126:515-523.

EEliflif FatmaFatma SSenen BW.inddBW.indd 4477 003-01-113-01-11 15:1715:17 23 Todd G, Dunlop K, McNaboe J, Ryan M, Carson D, Shields M. Growth and adrenal suppression in asthmatic children treated with high-dose fl uticasone propionate. . Lancet 1996;348:27-29. 24 Gulliver T, Eid N. Eff ects of glucocorticoids on the hypothalamic-pituitary-adrenal axis in children and adults. Immunol Allergy Clin North Am 2005;25:541-555. 25 Allen D. Eff ects of inhaled steroids on growth, bone metabolism, and adrenal function. Adv Pediatr 2006;53:101-110. 26 Cromer B, Scholes D, Berenson A, Cundy T, Clark M, Kaunitz A. Depot medroxyprogesterone acetate and bone mineral density in adolescents--the Black Box Warning: a Position Paper of the Society for Adolescent Medicine. . J Adolesc Health 2006;39:296-301. 27 Hartard M, Kleinmond C, Kirchbichler A, Jeschke D, Wiseman M, Weissenbacher E, et al. Age at fi rst oral contraceptive use as a major determinant of vertebral bone mass in female endurance athletes. Bone 2004;35:836-41. 28 Lloyd T, Lin H, Matthews A, Bentley C, Legro R. Oral contraceptive use by teenage women does not aff ect body composition. Obstet Gynecol 2002;100:235-9. 29 Heinemann L, Lewis M, Spitzer W, Thorogood M, Guggenmoos-Holzmann I, Bruppacher R. Throm- boembolic stroke in young women. A European case-control study on oral contraceptives. Trans- national Research Group on Oral Contraceptives and the Health of Young Women. . Contraception 1998;57:29-37. 30 Heinemann L, Lewis M, Thorogood M, Spitzer W, Guggenmoos-Holzmann I, Bruppacher R. Case- control study of oral contraceptives and risk of thromboembolic stroke: results from International Study on Oral Contraceptives and Health of Young Women. BMJ 1997;315:1502-4. 31 Lewis M. Myocardial infarction and stroke in young women: what is the impact of oral contraceptives? Am J Obstet Gynecol 1998;179:S68-77. 32 Samuelsson E, Hagg S, . Incidence of venous thromboembolism in young Swedish women and possibly preventable cases among combined oral contraceptive users. . Acta Obstet Gynecol Scand 2004;83::674-81. 33 Samuelsson E, Hedenmalm K, Persson I. Mortality from venous thromboembolism in young Swedish women and its relation to pregnancy and use of oral contraceptives--an approach to specifying rates. . Eur J Epidemiol 2005;20:509-16. 34 Rietveld R, ter Riet G, Bindels P, Schellevis F, van Weert H, . Do general practitioners adhere to the guideline on infectious conjunctivitis? Results of the Second Dutch National Survey of General Practice. BMC Fam Pract. 2007;8:54. 35 Rietveld R, ter Riet G, Bindels P, Bink D, Sloos J, van Weert H. The treatment of acute infectious conjunctivitis with fusidic acid: a randomised controlled trial.. Br J Gen Pract 2005;55:924-30. 36 Hamerlynck J, Rietveld R, Hooft L. From the Cochrane Library: Marginally higher chance of cure by antibiotic treatment in acute bacterial conjunctivitis. Ned Tijdschr Geneeskd. 2007;151:594-6. 37 Rose P, . Management strategies for acute infective conjunctivitis in primary care: a systematic review. Expert Opin Pharmacother 2007;12:1903-21. 38 Sharfstein J, North M, Serwint J. Over the counter but no longer under the radar--pediatric cough and cold medications. N Engl J Med 2007;357:2321-4. 39 Madsen H, Andersen M, Hallas J. Drug prescribing among Danish children: a population-based study. Eur J Clin Pharmacol 2001;57(59-65). 40 Schirm E, van den Berg P, Gebben H, Sauer P, al. e. Drug use of children in the community assessed through pharmacy dispensing data. Br J Clin Pharmacol 2000;50:473-8. 41 Pandolfi ni C, Bonati M. A literature review on off -label drug use in children. Eur J Pediatr 2005;164:552-8.

EEliflif FatmaFatma SSenen BW.inddBW.indd 4488 003-01-113-01-11 15:1715:17 Chapter 2.2: Drug utilisation in children-A cohort study in three European countries-

42 Madsen H, Andersen M, Hallas J. Drug prescribing among Danish children: a population-based study.. Eur J Clin Pharmacol 2001;57:159-65. 43 Clavenna A, Berti A, Gualandi L, Rossi E, De Rosa M, Bonati M. Drug utilisation profi le in the Italian pediatric population. Eur J Pediatr 2008;ON LINE. 44 Thrane N, Sørensen H. A one-year population-based study of drug prescriptions for Danish children. Acta Paediatr. 1999;88:1131-6. 45 Schirm E, van den Berg P, Gebben H, Sauer P, L. DJ-vdB. Drug use of children in the community assessed through pharmacy dispensing data.. Br J Clin Pharmacol 2000;50:473-8. 46 Silwer L, Lundborg C. Patterns of drug use during a 15 years period: data from a Swedish Country 1998-2002. Pharmacoepidemiol & Drug Safety 2005;14(11):813-820.

Copyright statement The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of all authors, an exclusive licence on a worldwide basis to the BMJ Publishing Group Ltd, and its Licensees to permit this article (if accepted) to be published in BMJ editions and any other BMJPGL products and to exploit all subsidiary rights.

Contributors All authors helped conceive the idea for the study, design the study, and analyze and interpret the data. MCJMS and KMCV drafted the manuscript. AN, EFS revised the manuscript. AC and ICKW supervised the study. MCJMS is guarantor.

Funding The work presented in this paper was funded by the European Community’s 6th Framework Programme. Project number LSHB-CT-2005-005216: TEDDY: Task force in Europe for Drug Development for the Young. The funding agency had no role in the collection of data, the analysis or interpretation of the data nor of the decision to submit.

Competing interests Miriam CJM Sturkenboom (received various unconditional research grants from pharmaceutical companies (Merck, Pfi zer, Johnson&Johnson, Amgen, Roche, Altana, GSK), MCJMS is consultant to Pfi zer, Celgene, Servier and Sanofi Aventis, none of issues is related to this topic), Katia MC Verhamme (no confl ict of interest), Alfredo Nicolosi ( AN has been reimbursed by Pfi zer Inc for attending several conferences; Pfi zer Inc has funded research to the National Research Council of Italy on sexual and bladder dysfunctions in adults) Macey L. Murray (no confl icts), Antje Neubert (no confl icts), Ian C. Wong (no confl icts), Daan Caudri (no confl icts), Gino Picelli (has conducted research funded by unconditional research grants in pediatrics from Merck USA, and BMS), Elif Fatma Sen (no confl icts), Carlo Giaquinto (Fee for speaking, consulting and research grants were provided by: Sanofi Pasteur, GSK, Abbott, BMS, Gilead, Abbott, Tibotec, Boheringer Ingelheim., GSK-Biologicals) Luigi Cantarutti (as president of SOSETE he has received research grants from GSK, Abbott, Merck and BMS), Paola Baiardi (no confl ict), Adriana Ceci (no confl ict)

Ethical approval: Internal review board of the integrated primary care information database. (IPCI) The use of IMS data for this study has been reviewed by an Independent Scientifi c and Ethics Committee

EEliflif FatmaFatma SSenen BW.inddBW.indd 4499 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 5500 003-01-113-01-11 15:1715:17 Chapter 2.3

Assessment of Pediatric asthma drug use in three European countries; a TEDDY study

E. Fatma Sen, Katia MC Verhamme, Antje Neubert, Yingfen Hsia, Macey Murray, Mariagrazia Felisi, Carlo Giaquinto, Geert W. ‘t Jong, Gino Picelli, Eugenio Baraldi, Alfredo Nicolosi, Adriana Ceci, Ian C. Wong, and Miriam CJM Sturkenboom, on behalf of the TEDDY European Network of Excellence.

Published in: European Journal of Pediatrics, 2010 Sep 2. [Epub ahead of print]

EEliflif FatmaFatma SSenen BW.inddBW.indd 5511 003-01-113-01-11 15:1715:17 What is already known · Asthma drugs are among the most prescribed drugs to children · Most frequently used asthma drugs are short-acting beta2-mimetics and inhaled steroids · There is a large variability in prescription rates between countries

What this study adds · Off -label use of asthma drugs in children is low · L inking multi-country databases makes it possible to study country specifi c pediatric drug use in a systematic manner without being hampered by methodological diff erences

EEliflif FatmaFatma SSenen BW.inddBW.indd 5522 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 5533 003-01-113-01-11 15:1715:17 Abstract

Asthma drugs are amongst the most frequently used drugs in childhood, but international comparisons on type and indication of use are lacking. The aim of this study was to describe asthma drug use in children with and without asthma in the Netherlands (NL), Italy (IT), and the United Kingdom (UK). We conducted a retrospective analysis of outpatient medical records of children 0-18 years from 1 January 2000 until 31 December 2005. For all children, prescription rates of asthma drugs were studied by country, age, asthma diagnosis and off -label status. One-year prevalence rates were calculated per 100 children per patient year (PY). The cohort consisted of 671,831 children of whom 49,442 had been diagnosed with asthma at any time during follow-up. ß2-mimetics and inhaled steroids were the most frequently prescribed asthma drug classes in NL (4.9 and 4.1/100 PY), the UK (8.7 and 5.3/100 PY) and IT (7.2 and 16.2/100 PY), respectively. Xanthines, anticholinergics, leukotriene receptor antagonists and anti-allergics were prescribed in less than 1 child per 100 per year. In patients without asthma, ß2-mimetics were used most frequently. Country diff erences were highest for steroids, (Italy highest), and for ß2-mimetics (the UK highest). Off -label use was low, and most pronounced for ß2-mimetics in children < 18 months (IT) and combined ß2-mimetics+anticholinergics in children <6 years (NL). Conclusion: This study shows that among all asthma drugs, ß2-mimetics and inhaled steroids are most often used, also in children without asthma, and with large variability between countries. Linking multi-country databases allows us to study country specifi c pediatric drug use in a systematic manner without being hampered by methodological diff erences. This study underlines the potency of healthcare databases in rapidly providing data on pediatric drug use and possibly safety.

EEliflif FatmaFatma SSenen BW.inddBW.indd 5544 003-01-113-01-11 15:1715:17 Chapter 2.3: Assessment of Pediatric asthma drug use in three European countries; a TEDDY study

Introduction

Respiratory drugs are amongst the most frequently prescribed drugs in children, and especially asthma drugs are frequently prescribed (36). Prevalence rates of asthma drug use range between 4 and 26 percent, depending on country, age range, and study period (Table 1) (8, 12, 16-18, 20, 22, 23, 25, 26, 41, 43). Asthma is defi ned as a chronic infl ammatory disorder of the airways and is associated with airway hyperresponsiveness that leads to recurrent episodes of wheezing, breathlessness, chest tightness, and coughing (1). In children aged 5 years and younger the clinical symptoms of asthma are variable and non-specifi c. Furthermore, routine testing for main asthma features is not possible in this age group, which hampers making a diagnosis. According to the Global Initiative for Asthma (GINA), bronchodilators, (inhaled short-acting ß2-mimetics: SABA; long-acting ß2-mimetics: LABA; anticholinergics: ACH), inhaled corticosteroids (ICS) and alternative (add-on) treatments such as leukotriene receptor antagonists (LTRA), xanthines and anti-allergics (cromones) all have a place in the treatment of childhood asthma, although at various stages depending on severity of illness (1). In pediatrics, asthma is one of the main indications for which these drugs are used. Apart from these respiratory drugs, systemic glucocorticoids are also used in the treatment of asthma. However, due to side eff ects such as potential adrenal suppression and growth impairment with long term use (1, 37), use of systemic glucocorticoids in asthmatic children is restricted to asthma exacerbations and very severe persistent asthma (1). Use of asthma drugs varies largely between countries as observed from the literature (Table 1). Part of this variability may be explained by diff erences in age distribution and type of data (i.e. surveys, prescriptions, dispensing), which hampers good comparisons between studies. To better explore diff erences in prescription rates of asthma drug use between countries, we conducted a retrospective cohort study in three European countries: the Netherlands, Italy and United Kingdom, using the same methodology and stratifying by age, sex and asthma diagnosis.

Methods

Data collection Data were obtained from general practice medical record databases in three countries according to a common protocol. Databases were the Pedianet database in Italy (29), the Integrated Primary Care Information (IPCI) database in the Netherlands (40), and the Mediplus Disease Analyzer-Mediplus database (IMS-DA) in the UK (42). Details are described elsewhere (36). Primary care in children is provided by the GP in the UK and the Netherlands, and by a family pediatrician in Italy. These primary care medical record databases were chosen because they

EEliflif FatmaFatma SSenen BW.inddBW.indd 5555 003-01-113-01-11 15:1715:17 Table 1: Overview of studies that have investigated the prevalence of asthma-related drug use in children Years # children Prevalence Author Type of data Country Age Study drugs studied in study estimate Pharmacy Bianchi et dispensing Italy 2003 55,242 0-<18 11.9 All asthma al (8) records (NHS drugs (R03) claims data) Pharmacy dispensing ß2-mim, Bollinger 57,586 – records USA 1997-2000 0-<18 10.4 – 13.2 ICS, LTRA, et al (12) 282,402 (Medicaid xanthines, OS claims data) Pharmacy Clavenna dispensing Italy 2006 923,353 0-<14 26 All asthma et al (16) records (NHS drugs (R03) claims data) Pharmacy All drugs, Clavenna dispensing Italy 2000 417,559 0-<14 22.2 including ATC et al (17) records (NHS R03 claims data) Pharmacy dispensing ß2-mim, records ICS, LTRA, Khaled et (Claims Canada 1999-2000 1,031,731 0-<18 18 xanthines, al (25) data from cromones, administrative ketotifen databases) Pharmacy ß2-mim, Korelitz et dispensing ICS, ACH, USA 2004-2005 4,259,103 0-<18 14.58 al (26) records (IL cromones, Claims data IL) xanthines, OS ß2-mim, ICS, Pharmacy cromones, Goodman 83,232 (in dispensing USA 1984-1993 0-<18 4.0 – 8.1 xanthines, et al (22) 93) records (GHC) fi xed combinations ß2-mim, ICS, LTRA, Pharmacy De Vries ACH, fi xed dispensing Netherlands 2002 73,416 0-<15 4.9 et al (18) combinations, records (IADB) antihistamines, OA Pharmacy Furu et al All asthma dispensing Norway 2004 1,192,841 0-<20 9.1 (20) drugs (R03) records (NPD) Pharmacy ß2-mim, ICS, Joesch et dispensing LTRA, ACH, USA 1996 6789 0-<18 6.5 al (23) records cromones, (MEPS) xanthenes

EEliflif FatmaFatma SSenen BW.inddBW.indd 5566 003-01-113-01-11 15:1715:17 Chapter 2.3: Assessment of Pediatric asthma drug use in three European countries; a TEDDY study

Table 1: Continue Years # children Prevalence Author Type of data Country Age Study drugs studied in study estimate ß2-mim, ICS, Survey, claims, Wang et al LTRA, ACH, prescriptions USA ‘96+’98+2000 15,554 5-17 5.2 (41) cromones, (MEPS) xanthines, OS ß2-mim,ICS, Zuidgeest Prescription Netherlands 2001 74,580 0-<18 7.5 cromones, et al (43) data (GP) LTRA Legend: Study drugs: ß2-mim = ß2-mimetics, ICS = inhaled corticosteroids, LTRA = leukotriene-receptor antagonists, ACH = anticholinergics ATC R03 = drugs for obstructive airway diseases, OS = oral steroids, OA = oral adrenergics Type of data: NHS = national health service, NPD = Norwegian Prescription Database, GHC = Group Health Cooperative of Puget Sound, MEPS = Medical Expenditur Panel Survey, IL = Ingenix LabRx, IADB = InterAction DataBase

contain detailed information on the population, diagnoses and prescriptions. The geographic location of the databases allowed us to study North and South European diff erences in drug prescribing.

Study population The study population consisted of all children aged 0-18 years (0-11 years in Italy) during the study period 1 January 2000– 31 December 2005 (1 January 2001– 31 December 2005 for Italy). All persons needed to have a database history of ≥ 6 months, except for children born during the study period. For each child the person years of follow-up were calculated, stratifi ed by calendar year, age group, sex, and asthma diagnosis. For privacy issues, some databases only contain the year of birth, therefore age was assessed at the 1st of January of each year, and in order to identify infants, preschoolers and adolescents separately. We used modifi ed ICH (the International Conference of Harmonization) age categories to stratify by age (<2, >2-<4, 5-11 and >12 years of age) (2). Each child was followed from study start until the patient left practice or end of study period (whichever was earliest).

Assessment of asthma For each child we assessed whether a diagnosis of asthma was made at anytime in its medical records using the following diagnostic codes: ICPC code R96 in IPCI, ICD-9 code 493 in Pedianet and ICD-10 code J45 in IMS and free text (4, 5, 27). If at any time during the follow-up asthma was diagnosed, the child was considered asthmatic for the entire perio

Classifi cation of prescriptions Drug prescriptions were categorized based on the World Health Organization’s Anatomical Therapeutic Chemical (ATC) classifi cation system (6). All prescriptions for “obstructive airway

EEliflif FatmaFatma SSenen BW.inddBW.indd 5577 003-01-113-01-11 15:1715:17 Table 2: Population size by country, age, sex and asthma diagnosis NL Italy$ UK Total Patient Children Total Children Total Children Total Children characteristics with asthma pop. with asthma pop. with asthma pop. with asthma No.# (% of No.# (% of No.# (% of No.# (% of total) total) total) total) Age <2 years 2,165 (8.4) 25.694 2,539 (4.6) 54,999 2,977 (3.8) 79,246 7,681 (5.7) ≥2-≤4 years 2,820 (11.8) 23,902 4,269 (7.3) 58,735 7,610 (7.5) 101,835 14,699 (8.0) 5-≤11 years 5,545 (11.7) 47,435 5,399 (7.5) 72,264 19,856 (12.8) 155,488 30,800 (11.2) 12-≤18 years 3,689 (8.3) 44,636 19,809 (13.3) 149,093 23,498 (12.1) Sex Females 3,940 (7.9) 49,709 2,851 (4.7) 60,089 14,377 (6.5) 219,669 21,167 (6.4) Males 5,419 (10.5) 51,850 4,730 (7.2) 65,361 18,125 (8.1) 225,153 28,274 (8.3) Total 9,359 (9.2) 101,559 7,581 (6.0) 125,450 32,502 (7.9) 444,822 49,442 (7.4) PY=person-years; #The number of children in the various age groups does not add up to the total since one child can contribute to more than one age category during the study period. $ For Italy the age range included only 0-11 years for the years 2001-2005

diseases” (ATC R03) and “glucocorticosteroids for systemic use” (ATC H02AB) were obtained from the prescription fi les and further categorized into: ß2-mimetics (either short-acting [SABA] or long-acting [LABA]), inhalation glucocorticosteroids, inhalation anticholinergics, fi xed combination of SABA + short-acting anticholinergics, fi xed combination of LABA + inhalation glucocorticosteroids, anti-allergic agents (best known as cromones), xanthines and leukotriene receptor antagonists (LTRA). Use of ß2-mimetics+ICS was classifi ed as a non-fi xed combination if the child used ß2-mimetics and an ICS in the same year. For each drug class, we calculated prevalence rates for the most frequently prescribed drugs (top 3) by country. Due to diff erences in labeling status between countries, off -label use was assessed per country. A drug was classifi ed as being “off -label” if the child receiving the prescription had an age below the minimum age as specifi ed in the summary product characteristics (SPC) or the drug was given for a diff erent indication than specifi ed in the SPC. This is in line with the new defi nition of pediatric off -label use provided by the Task Force in Europe for Drug development for Young (31).

Statistical analysis Annual prevalence rates (per 1000 PY or per 100 PY) of asthma drug use were calculated by counting the number of children using each drug or drug class by calendar year, age, sex, asthma status and country and dividing this number by the number of person years attributed by the diff erent strata. User prevalence rates can be interpreted as the number of children per 1,000 children that uses a specifi c drug (class) during one year. If expressed per 100 PY it can be freely translated as the percentage of children using these drugs in one year. The fact that we used annual prevalence rates means that a child who uses one drug (class) multiple times in one year, will be counted only once for that year. The use of person-years rather than persons in

EEliflif FatmaFatma SSenen BW.inddBW.indd 5588 003-01-113-01-11 15:1715:17 Chapter 2.3: Assessment of Pediatric asthma drug use in three European countries; a TEDDY study

the denominator was necessary due to the dynamic nature of the cohort where persons could have variables durations of follow-up. Diff erences in prescription rates between countries were tested by means of the X2-test.

Results

Patient characteristics Our population consisted of 671,831 children (66% UK, 19% IT, 15% NL) of whom 49,442 (7.4%) had asthma at any point during follow-up.

Asthma drug classes The user prevalence rate of any asthma drug was highest in young children and decreased with age (Figure 1), it was highest in Italy for each age category. The diff erence between NL, the UK and IT is largest for the youngest age categories with user prevalence rates of 25-30/100 PY in Italy, rates of 12 in the UK, and 10/100 PY in NL (Figure 3-4). In the youngest age groups, prescription rates were higher for boys. This pattern switched in adolescence (between the ages 14 to 16), where user prevalence rates for almost all drug classes were higher in girls (Figure 1). ß2-mimetics and ICS were each used by 1-10 children/100 PY in all age categories in each country (Figure 2). In Italy, prevalence of use of ICS was higher than use of ß2-mimetics and even exceeded 10 /100 PY in the youngest age group. The fi xed combinations of anticholinergics+SABA were frequently used in the Netherlands and Italy but not in the UK,

boys - Ipci girls - Ipci boys - IMS girls - IMS boys - Pedianet girls - Pedianet

350

300

250

200

150

users/1000 PY users/1000 100

0 < 1 1 2 3 4 5 6 7 8 9 101112131415161718 year age Figure 1: Age, country and sex specifi c user prevalence rates of asthma drugs (all ATC R03)

EEliflif FatmaFatma SSenen BW.inddBW.indd 5959 003-01-113-01-11 15:1715:17 250

200

150

100

50 number of of number PY users/1000

0 NL IT UK NL IT UK NL IT UK NL IT UK NL IT UK NL IT UK NL IT UK NL IT UK

ß2-mimetics ß2-mimetics (single ICS (alone) ACH (alone) SABA+ACH (fixed) LABA+ICS (fixed) ß2+ICS (non fixed) Systemic use) corticosteroids

asthma drug class

<2 years 2-4 years 5-11 years 12-18 years Figure 2: Annual user prevalence rates of the most frequently prescribed respiratory drugs NL: the Netherlands, IT: Italy, UK: the United Kingdom, ß2-mimetics: inhaled ß2-mimetics (this can be single or non-fi xed combination use), ICS: inhalation corticosteroids, ACH: anticholinergics, SABA+ACH: fi xed combinations of short-acting ß2-mimetics+anticholinergics, LABA+ICS: fi xed combinations of long- acting ß2-mimetics+inhalation corticosteroids, ß2+ICS(non fi xed): non-fi xed combination of ß2-mimetics (LABA or SABA)+inhalation corticosteroids

children with asthma

700

600

500

400

300

200

100 numbet of users/1000 PY users/1000 of numbet

0 NL IT UK NL IT UK NL IT UK NL IT UK NL IT UK NL IT UK

ß2-mimetics ICS ACH SABA+ACH (fixed) LABA+ICS (fixed) Systemic corticosteroids

asthma drug class 0-<2 years 2-4 years >4-11 years >11-18 years Figure 3: Prevalence of respiratory drug use by age category and country in children with asthma b2-mim= ß2-mimetics; ics=inhalation corticosteroids; ach=anticholinergics; saba+ach= fi xed combination of short-acting ß2-mimetics+anticholinergics; laba+ics= fi xed combination of long-acting ß2- mimetics+inhalation corticosteroids; scort=systemic corticosteroids

where mostly anticholinergics were used as single compound. Leukotriene receptor antagonists, anti-allergics and xanthines were used by less than 1 child/100 PY in each country and each age category, which is in concordance with protocols concerning asthma treatment. Fixed combinations of ICS+LABA were infrequently used in young children but exceeded 1/100 PY in adolescents in the Netherlands. Systemic use of steroids was very heterogeneous between

EEliflif FatmaFatma SSenen BW.inddBW.indd 6600 003-01-113-01-11 15:1715:17 Chapter 2.3: Assessment of Pediatric asthma drug use in three European countries; a TEDDY study

children without asthma 250

200

150

100

50 numbet of users/1000 PY of users/1000 numbet

0 NL IT UK NL IT UK NL IT UK NL IT UK NL IT UK NL IT UK

ß2-mimetics ICS ACH SABA+ACH (fixed) LABA+ICS (fixed) Systemic corticosteroids asthma drug class 0-<2 years 2-4 years >4-11 years >11-18 years Figure 4: Prevalence of respiratory drug use by age category and country in children without asthma b2-mim= ß2-mimetics; ics=inhalation corticosteroids; ach=anticholinergics; saba+ach= fi xed combination of short-acting ß2-mimetics+anticholinergics; laba+ics= fi xed combination of long-acting ß2- mimetics+inhalation corticosteroids; scort=systemic corticosteroids

countries. User rates exceeding 10/100 PY were observed in Italy, especially in the youngest children, whereas user rates were 5-8 folds lower in the UK and almost 20 fold lower in the Netherlands.

User prevalence rates for individual drugs The prevalence of SABA use in children with asthma was similar in the Netherlands and Italy but much higher in the UK (30 vs. 60/100 PY). In children without asthma diagnosis the prevalence was between 1 and 8/100 PY. The most frequently prescribed SABAs were salbutamol and terbutaline in NL and the UK, whereas terbutaline was not frequently used in Italy (Table 3). User rates of LABA (this can be combined or single use) were very low (< 1/100 PY), but strongly increased with age in children with asthma. Most frequently used drugs were salmeterol and formoterol. In children without asthma, use of LABA was negligible (Table 3 and Figure 3). Single use of ß2-mimetics (without a prescription for ICS in the same year) was highest in the UK (195/1000 PY in children with asthma and 23 in children without). In NL and IT prevalence rates were 127 and 75/1000 PY in children with asthma and 12 and 18/1000 PY in children without an asthma diagnosis. In both countries prevalence rates decreased gradually with increasing age. ICS were most frequently prescribed in Italy, especially in children < 2 years of age. In children with asthma, ICS use was highest in the UK. The choice of drug diff ered; fl uticasone was most often prescribed in NL and beclomethasone in the UK and Italy. Among the children with asthma who used ß2-mimetics, 42% were not prescribed ICS in that year in NL, 26% in Italy, and 33% in the UK. In children without asthma, ICS use was low in NL and the UK but reached rates of more than 10/100 PY in Italy. Oral corticosteroid use was highest in Italy, with prescription rates of 26 /100 PY in children with asthma and 7 /100 PY in children without asthma. Rates were especially high in the younger age categories. In the other countries, prescription rates in children with asthma were much

EEliflif FatmaFatma SSenen BW.inddBW.indd 6611 003-01-113-01-11 15:1715:17 Table 3: Prevalence of respiratory drug use per country by treatment class and asthma diagnosis Netherlands Italy United Kingdom no no no asthma asthma asthma asthma asthma asthma SABA Salbutamol 276 19 Salbutamol 283 59 Salbutamol 519 32 Terbutaline 39 2 Terbutaline 1.4 0.1 Terbutaline 88 2.8 Overall 293 17 284 59 596 35 LABA Salmeterol 14 0.4 Salmeterol 13 0.5 Salmeterol 40 0.6 Formoterol 7.6 0.2 Formoterol 1.6 0.1 Formoterol 2.5 0.02 Overall 21 0.6 15 0.6 42 0.6 ICS Fluticasone 162 8 Beclomethasone 239 98 Beclomethasone 313 11 Budesonide 68 4 Fluticasone 122 10 Budesonide 72 1.9 Beclomethasone 50 2 Flunisolide 76 32 Fluticasone 63 1.5 Overall 271 14 404 143 437 14 LTRA Montelukast 5.5 0.1 Montelukast 37 1.7 Montelukast 15 0.3 Zafi rlukast 0.0 0.0 Zafi rlukast 0.2 0.02 Overall 5.5 0.1 38 1.7 15 0.3 Short-acting anticholinergics Ipratropium 15.7 1.6 Ipratropium 6.2 2.4 Ipratropium 9.1 2.7 bromide Oxitropium 0.3 0.1 Oxitropium 0.04 0 Overall 16 1.6 6.4 2.5 9.1 2.7 ß2-agonists+ICS (non fi xed) 177 5 217 42 404 12 SABA+SACH (fi xed) Fenoterol+ Salbutamol+ Salbutamol+ 19 4.1 104 25 0.7 0.06 Ipratropium Ipratropium Ipratropium Salbutamol+ Fenoterol+ Fenoterol+ 0.9 0.9 0.1 0.0 0.05 0 Ipratropium Ipratropium Ipratropium Overall 20 4.4 104 25 0.07 0 LABA+ICS (fi xed) Salmeterol+ Salmeterol+ Salmeterol+ 33 1.3 37 1.3 31 0.5 Fluticasone Fluticasone Fluticasone Formoterol+ Formoterol+ Formoterol+ 6.8 0.3 2.1 0.1 6.1 0.07 Budesonide Budesonide Budesonide overall 60 2.3 38 1.4 37 0.6 Asthma = prevalence in children with asthma; no asthma = prevalence in children without asthma; SABA= short-acting ß2-mimetics; LABA= long-acting ß2-mimetics; ICS=inhalation corticosteroids; LTRA= leukotriene-receptor antagonists; ß2-agonists+ICS (non fi xed) = non-fi xed combination of ß2-mimetics+inhalation corticosteroids; SABA+SACH (fi xed) = fi xed combination of short-acting ß2-mimetics+short-acting anticholinergics; LABA+ICS (fi xed) = fi xed combination of long-acting ß2- mimetics+inhalation corticosteroids. * Could either be the combination of LABA+ICS are SABA+ICS but always a loose combination (more than 1 device)

EEliflif FatmaFatma SSenen BW.inddBW.indd 6622 003-01-113-01-11 15:1715:17 Chapter 2.3: Assessment of Pediatric asthma drug use in three European countries; a TEDDY study

lower in each age category, both in children with and without asthma. The most frequently prescribed drugs were betamethasone in Italy and prednisolone in NL and the UK.

Combination preparations Use of the fi xed combinations of short acting SABA+ACH was low overall, but most frequent in Italy. Prescription rates decreased with age. For the fi xed combinations of ICS+LABA, use was highest in NL with 6/100 PY in children with asthma. A clear increase of prescription rates with age was seen in all 3 countries. Use of the loose combination of ß2-mimetics+ICS was highest in IT (55/1000 PY), followed by the UK (48/1000 PY) and NL (24/1000 PY). In children with asthma prevalence rates were highest in the UK (404/1000 PY), followed by IT (217) and NL (177)

Off -label use Off -label use defi ned by use below the minimum age as specifi ed in the SPC was investigated in Italy and NL. In Italy, salbutamol is indicated for use in children from the age of 18 months onwards, however user rates were 22.7 /100 PY in children with asthma and 7.5 in children without asthma diagnosis below 18 months. In NL off -label use was observed for the fi xed combinations of ß2-mimetics+anticholinergics. The minimum age for use of these drugs is 6 years in NL, however, prescription rates in children <6 years were 0.5-3.5 /100 PY in NL (data not shown). Off -label use defi ned as use for other indications than specifi ed in the SPC (for drugs that only had asthma or COPD as indication) in children without recorded asthma diagnosis was observed in all 3 countries for several drugs (Table 4). In Italy budesonide had the highest off - label prescription rate, in NL salbutamol, and in the UK beclomathesone.

Discussion

This paper provides detailed information on asthma drug use in three European countries, and clearly demonstrates the similarities and diff erences in prescription rates. Whereas short-acting ß2-mimetics are the most frequently used drugs for treatment of asthma in the Netherlands and the UK, inhaled and systemic corticosteroids are most frequently used in Italy, especially in the youngest age groups. Long-acting products, anticholinergics, xanthines, and LTRAs were in general infrequently prescribed in all countries. Our data are in line with previous studies on asthma drug use in children, which also demonstrated that short-acting ß2-mimetics and inhaled corticosteroids were prescribed most, but these studies often considered only one country or a region (20, 21), whereas our study captures multiple countries in diff erent European regions. Our data also confi rm fi ndings of a recent survey showing high use of inhaled ß2-mimetics in northern Europe and high usage

EEliflif FatmaFatma SSenen BW.inddBW.indd 6633 003-01-113-01-11 15:1715:17 Rate* Prescription - label % off Drug Rate* Prescription - label % off ed in the Summary Characteristics of Product in the table listed -label use are c drug c Drug Rate* Prescription Only those with off -label status. - -label use of the specifi label % off other indications than specifi ned as use for Netherlands Italy Kingdom United -label use defi Off of off -label = the percentage SalbutamolTerbutalineSalmeterolFormoterolBeclomethasone 32.8Fenoterol+Ipratropium 30.7Salbutamol+IpratropiumSalmeterol+Fluticasone 19.0 63.0 15 27.2 72.5Formoterol+Budesonide 16.4 2.1 acidCromoglicic 24.4 Budesonide 0.4 4.1 28.4Montelukast 2.3 0.3 Salmeterol+Fluticasone 0.2 Formoterol+Budesonide 1.3 Montelukast Nedocromil 0.3 acid Cromoglicic 31.6 26.3 29.1 80.0 16.0 1.3 0.1 0.1 51.3 36.1 14.8 Salmeterol 54.5 0.1 Beclomethasone Terbutaline 0.3 1.7 2.1 Budesonide Ipratropium Fluticasone 25.8 12.1 23.5 11 0.6 21.0 74.2 2.7 Salbutamol+Ipratropium 18.7 Formoterol+Budesonide 1.9 45.5 acid Cromoglicic 2.7 1.5 10.8 0.1 0.1 34.7 0.2 Drug -*prescription rate = the prescription rate of the drug /1000 PY in children without asthma in children of the drug /1000 PY rate = the prescription rate -*prescription asthma a diagnosis Budesonide in Italy receiving do not have 80% of the children -e.g.: off checked for -all study drugs were Table 4: Table -% off

EEliflif FatmaFatma SSenen BW.inddBW.indd 6644 003-01-113-01-11 15:1715:17 Chapter 2.3: Assessment of Pediatric asthma drug use in three European countries; a TEDDY study

of corticosteroids in southern Europe (11). The north south gradient that we observed in the prescription of asthma drugs can not be explained by country specifi c diff erences in asthma treatment protocols, as the protocols in the Netherlands; the UK and Italy are all based on the GINA guidelines. There are various prescribing behaviors which deserve specifi c attention since they may be associated with a diff erential risk of adverse drug reactions. First of all, SABA use is high in the lower age categories which may at least partly be explained by the fact that they are recommended on an as-needed basis for acute wheezing in preschool children (13). However, paradoxical responses like deterioration of asthma with use of these drugs have been reported in infants (38). Secondly, prescription rates for long-acting ß2-mimetics were low for all 3 countries, which is in line with guidelines and recommendable since a black box warning was issued by the FDA due to safety issues in children, especially in mono-use (19). Thirdly, in children with asthma the percentage of children using β2-mimetics without ICS as controller was substantial, especially in the UK and Netherlands. A possible explanation could be that these children were intermit- tently treated with SABA’s on an as-needed basis as defi ned in the 1st step of the guidelines (1). However, chronic use of ß2-mimetics without ICS is contraindicated in children with asthma because it may lead to increased airway infl ammation and worsening of asthma control, and is in severe cases even associated with death (28, 33). Fourth, use of ICS was very high in children under the age of 5 in Italy; this level of use certainly raises a question about potential overprescription, especially in children without asthma. How- ever, a recent survey among children 1-5 years of age showed that southern Europe had the highest rate of wheezing symptoms (48%) compared with northern Europe (29%), which may explain some of the high use of ICS (11). Although there is some evidence supporting the eff ectiveness of regular inhaled corticosteroids in preschool wheeze (9) its use remains controversial (13, 24). Long term use of steroids (inhaled and specifi cally oral) might cause side eff ects such as temporary growth retardation and adrenal suppression (1, 7). Less severe but more frequent and relevant side eff ects in children treated with ICS are oropharyngeal candidiasis, dysphonia, refl ex cough and pharyngitis (14). Fifth, systemic corticosteroids are not the fi rst choice of treatment in asthma and should only be used for exacerbations. A recent study showed that oral prednisolon and placebo did not diff er in duration of hospitalization in children with mild to moderate wheezing associated with a viral infection (32). Although the high use of oral steroids in Italy is consistent with previous study fi ndings (1, 8, 15), it may be slightly overestimated. Betamethasone (dispensed as capsules) is often administered via a nebulizer meaning that the actual oral use will be lower than reported, unfortunately we could not distinguish between these modes of administration. Fixed combination therapies of an ICS+LABA are relatively new in the treatment of childhood asthma and their long term safety is not well known. Use was highest in NL, similar to a recently reported increase in use of the fi xed combination of ICS+LABA in Denmark (10). Although fi xed

EEliflif FatmaFatma SSenen BW.inddBW.indd 6655 003-01-113-01-11 15:1715:17 combination products of ICS+LABA are only licensed for use in children over 4 years, these drugs were also prescribed to children <2 years in the Netherlands. We found only 1 small retrospective study investigating the eff ect of a fi xed combination of ICS+LABA in children <4 years, reporting a reduction in morbidity and an acceptable safety profi le (34). So far, the safety and effi cacy data of the fi xed combination of ICS+LABA in children <2 years is too scarce to justify prescribing these drugs in young children. In general, off -label use of asthma drugs defi ned as use below the minimum age as reported in the SPCs was low except for the fi xed combination of ß2-mimetics+anticholinergics in children <6 years in the Netherlands and the use of SABA in children <18 months in Italy. Prescription rates for asthma drugs in children without an asthma diagnosis are much lower than in children diagnosed with asthma but still surprisingly high, especially for ICS and SABA. These drugs may have been used to treat wheezy young children with other respiratory conditions, such as respiratory tract infections (e.g. bronchiolitis) or fi rst symptoms of undiagnosed asthma in young preschool children, but still, overuse cannot be excluded. Use of asthma drugs is high, although very few long-term safety studies are available for children. The Pediatric Committee of the European Medicines Agency has recently published a pediatric needs list, containing drugs that are being used in children and where information on pharmacokinetics, effi cacy and safety is urgently needed (3). This list includes frequently used respiratory drugs such as salbutamol, fl uticasone, montelukast and many others. Although RCTs are the gold standard for effi cacy assessment, safety studies are more eff ectively conducted through large scale postmarketing observational studies (35). Clinical trials are often small and too short for safety assessment. Pharmacoepidemiological studies are a valid tool for assessing safety under real life circumstances and often rely on secondary use of routine care data from claims databases and medical records. This study, as well as a previous inventory on databases with pediatric data, shows that exposure and outcome information is available on a large pediatric population (30, 36). Pharmacoepidemiological studies should be promoted to improve the safety assessment of pediatric drugs since RTCs are hampered by methodological diffi culties, especially in the pediatric subpopulation, and consequently, results may not be prompt and suffi cient enough for safety assessment. As for all observational research, our data has some limitations. First of all, as we used primary care data, we may not capture all specialist prescriptions in the UK and the Netherlands. This will minimally infl uence our results since asthma is a condition often dealt with in primary care, and asthma drugs originally prescribed by a specialist are often continued or prescribed by general practitioners (39). Also, the health care provider in the studied countries diff ered: GP data from the UK and NL and family pediatrician data from IT. However, we do not believe that this would hinder comparability between databases as the nature of the databases is the same. If there are diff erences between countries, this refl ects diff erences in primary care prescribing behavior and not in the type of data.

EEliflif FatmaFatma SSenen BW.inddBW.indd 6666 003-01-113-01-11 15:1715:17 Chapter 2.3: Assessment of Pediatric asthma drug use in three European countries; a TEDDY study

Furthermore, it is not possible to diagnose asthma in young pre-school children and the diagnosis of asthma is often made at a later age, which could lead to misclassifi cation of the indication in young children. However, to minimize the possibility of misclassifi cation, we searched the entire medical record of a child for the diagnosis of asthma. If the diagnosis was made at a later age, the child was counted as having asthma during the whole study period. To conclude, our results show high use of SABA and ICS, although not always combined, and low use of other asthma drugs. Safety data on specifi c asthma drugs are needed according to the European Medicines Agency (3). This study underlines the potency of healthcare databases in rapidly providing data on pediatric drug use and possibly safety.

EEliflif FatmaFatma SSenen BW.inddBW.indd 6677 003-01-113-01-11 15:1715:17 Acknowledgements: We would like to thank Peter Stephens of IMS Health for providing the IMS-DA database. We thank all of the physicians contributing data to the PEDIANET, IPCI and IMS-DA databases.

None of the authors has a confl ict of interest. The principal investigator had full access to all of the data in the study and takes responsibility for their integrity and the accuracy of the data analysis.

Funding: Funded by the European Community’s 6th Framework Programme. Project number LSHB-CT-2005-005216: TEDDY: Task force in Europe for Drug Development for the Young.

EEliflif FatmaFatma SSenen BW.inddBW.indd 6688 003-01-113-01-11 15:1715:17 Chapter 2.3: Assessment of Pediatric asthma drug use in three European countries; a TEDDY study

References

1 The Global Strategy for Asthma Management and Prevention, Global Initiative for Asthma (GINA) 2007.; Available from: http://www.ginasthma.org. 2 ICH Guidelines. Available from: http://www.ich.org/cache/compo/276-254-1.html. 3 List of paediatric needs (as established by the Paediatric Working Party). Available from: http://www. emea.europa.eu/htms/human/paediatrics/inventory.htm. 4 US Centers for Disease Control and Prevention. International Classifi cation of Diseases, ninth revision, clinical modifi cation (ICD-9-CM). Available from http://www.cdc.gov/nchs/icd/icd9cm.htm. Access date: 5 February 2010. 5 US Centers for Disease Control and Prevention. International Classifi cation of Diseases, Tenth Revision (ICD - 10). [cited 2010 16 June]; Available from: http://www.cdc.gov/nchs/icd/icd10.htm. 6 WHO Collaborating Center for Drug Satistics Methodology. Available from: http://www.whocc.no/ atcddd/. 7 Bacharier LB, Boner A, Carlsen KH, et al. 2008 Diagnosis and treatment of asthma in childhood: a PRACTALL consensus report. Allergy. 2008 Jan;63(1):5-34. 8 Bianchi M, Clavenna A, Labate L, et al. 2009 Anti-asthmatic drug prescriptions to an Italian paedriatic population. Pediatr Allergy Immunol. 2009 Sep;20(6):585-91. 9 Bisgaard H, Gillies J, Groenewald M, et al. 1999 The eff ect of inhaled fl uticasone propionate in the treatment of young asthmatic children: a dose comparison study. American journal of respiratory and critical care medicine. 1999 Jul;160(1):126-31. 10 Bisgaard H, Szefl er S. 2006 Long-acting beta2 agonists and paediatric asthma. Lancet. 2006 Jan 28;367(9507):286-8. 11 Bisgaard H, Szefl er S. 2007 Prevalence of asthma-like symptoms in young children. Pediatric pulmonology. 2007 Aug;42(8):723-8. 12 Bollinger ME, Smith SW, LoCasale R, et al. 2007 Transition to managed care impacts health care service utilization by children insured by Medicaid. J Asthma. 2007 Nov;44(9):717-22. 13 Brand PL, Baraldi E, Bisgaard H, et al. 2008 Defi nition, assessment and treatment of wheezing disorders in preschool children: an evidence-based approach. Eur Respir J. 2008 Oct;32(4):1096-110. 14 Buhl R. 2006 Local oropharyngeal side eff ects of inhaled corticosteroids in patients with asthma. Allergy. 2006 May;61(5):518-26. 15 Clavenna A, Berti A, Gualandi L, et al. 2008 Drug utilisation profi le in the Italian paediatric population. European journal of pediatrics. 2008 Apr 30. 16 Clavenna A, Berti A, Gualandi L, et al. 2009 Drug utilisation profi le in the Italian paediatric population. European journal of pediatrics. 2009 Feb;168(2):173-80. 17 Clavenna A, Rossi E, Berti A, et al. 2003 Inappropriate use of anti-asthmatic drugs in the Italian paediatric population. European journal of clinical pharmacology. 2003 Oct;59(7):565-9. 18 de Vries TW, Tobi H, Schirm E, et al. 2006 The gap between evidence-based medicine and daily practice in the management of paediatric asthma. A pharmacy-based population study from The Netherlands. European journal of clinical pharmacology. 2006 Jan;62(1):51-5. 19 FDA. FDA Public Health Advisory. Available from: http://www.fda.gov/cder/drug/advisory/LABA.htm. 20 Furu K, Skurtveit S, Langhammer A, et al. 2007 Use of anti-asthmatic medications as a proxy for prevalence of asthma in children and adolescents in Norway: a nationwide prescription database analysis. European journal of clinical pharmacology. 2007 Jul;63(7):693-8. 21 Gislason T, Olafsson O, Sigvaldason A. 1997 Users of antiasthma drugs in Iceland: a drug utilization study. Eur Respir J. 1997 Jun;10(6):1230-4.

EEliflif FatmaFatma SSenen BW.inddBW.indd 6699 003-01-113-01-11 15:1715:17 22 Goodman DC, Lozano P, Stukel TA, et al. 1999 Has asthma medication use in children become more frequent, more appropriate, or both? Pediatrics. 1999 Aug;104(2 Pt 1):187-94. 23 Joesch JM, Kim H, Kieckhefer GM, et al. 2006 Does your child have asthma? Filled prescriptions and household report of child asthma. J Pediatr Health Care. 2006 Nov-Dec;20(6):374-83. 24 Kaditis AG, Winnie G, Syrogiannopoulos GA. 2007 Anti-infl ammatory pharmacotherapy for wheezing in preschool children. Pediatric pulmonology. 2007 May;42(5):407-20. 25 Khaled LA AF, Brogan T, Fearnley J, Grahan J, MacLeod S, McCormick J. 2003 Prescription medicine use by one million Canadian children. Paediatr Child Health 2003, VOL 8; SUPP/A. 2003. 26 Korelitz JJ, Zito JM, Gavin NI, et al. 2008 Asthma-related medication use among children in the United States. Ann Allergy Asthma Immunol. 2008 Mar;100(3):222-9. 27 Lamberts H WM, Hofmans-Okkes I:. 1993:242. The International Classifi cation of Primary Care in the European Community. Oxford University Press;. 1993:242. 28 Lanes SF, Garcia Rodriguez LA, Huerta C. 2002 Respiratory medications and risk of asthma death. Thorax. 2002 Aug;57(8):683-6. 29 Menniti-Ippolito G, Raschetti R, Da Cas R, et al. 2000 Active monitoring of adverse drug reactions in children. Italian Paediatric Pharmacosurveillance Multicenter Group. Lancet. 2000;355:1613-4. 30 Neubert A, Sturkenboom MC, Murray ML, et al. 2008 Databases for pediatric medicine research in Europe--assessment and critical appraisal. Pharmacoepidemiology and drug safety. 2008 Dec;17(12):1155-67. 31 Neubert A. BA, Catapano M., Baiardi P, Guiaquinto C., Knibbe C. Sturkenboom M CJM, Wong I CK, Ceci A. On behalf of the TEDDY Network of Excellence. Defi ning Off -label and Unlicensed Use of Medicines for Children: Results of a Delphi survey. (Pharmacol Res., In Press). 32 Panickar J, Lakhanpaul M, Lambert PC, et al. 2009 Oral Prednisolone for Preschool Children with Acute Virus-Induced Wheezing. The New England journal of medicine. 2009 Jan 22;360(4):329-38. 33 Salpeter SR, Ormiston TM, Salpeter EE. 2004 Meta-analysis: respiratory tolerance to regular beta2- agonist use in patients with asthma. Annals of internal medicine. 2004 May 18;140(10):802-13. 34 Sekhsaria S, Alam M, Sait T, et al. 2004 Effi cacy and safety of inhaled corticosteroids in combination with a long-acting beta2-agonist in asthmatic children under age 5. J Asthma. 2004 Aug;41(5):575-82. 35 Sorensen HT, Lash TL, Rothman KJ. 2006 Beyond randomized controlled trials: a critical comparison of trials with nonrandomized studies. Hepatology (Baltimore, Md. 2006 Nov;44(5):1075-82. 36 Sturkenboom MC, Verhamme KM, Nicolosi A, et al. 2008 Drug use in children: cohort study in three European countries. BMJ (Clinical research ed. 2008;337:a2245. 37 Szefl er SJ. 1991 Glucocorticoid therapy for asthma: clinical pharmacology. The Journal of allergy and clinical immunology. 1991 Aug;88(2):147-65. 38 Taylor DR, Sears MR, Cockcroft DW. 1996 The beta-agonist controversy. The Medical clinics of North America. 1996 Jul;80(4):719-48. 39 van der Lei J, Duisterhout J, Westerhof H, et al. 1993 The introduction of computer-based patient records in the Netherlands. Ann Intern Med. 1993;119(10):1036-41. 40 Vlug AE, van der Lei J, Mosseveld BM, et al. 1999 Postmarketing surveillance based on electronic patient records: the IPCI project. Methods of information in medicine. 1999 Dec;38(4-5):339-44. 41 Wang LY, Zhong Y, Wheeler L. 2006 Asthma medication use in school-aged children. J Asthma. 2006 Sep;43(7):495-9. 42 Wong I, Murray M. 2005 The potential of UK clinical databases in enhancing paediatric medication research. Br J Clin Pharmacol. 2005;59(6):750-5.

EEliflif FatmaFatma SSenen BW.inddBW.indd 7700 003-01-113-01-11 15:1715:17 Chapter 2.3: Assessment of Pediatric asthma drug use in three European countries; a TEDDY study

43 Zuidgeest MG, van Dijk L, Smit HA, et al. 2008 Prescription of respiratory medication without an asthma diagnosis in children: a population based study. BMC health services research. 2008;8:16.

EEliflif FatmaFatma SSenen BW.inddBW.indd 7711 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 7722 003-01-113-01-11 15:1715:17 Cha pter 2.4

Eff ects of safety warnings on prescription rates of cough and cold medicines in children below 2 years of age

E. Fatma. Sen, Katia MC Verhamme, Mariagrazia Felisi, Geert W. ‘t Jong, Carlo Giaquinto, Gino Picelli, Adriana Ceci, and Miriam CJM Sturkenboom, on behalf of the TEDDY European Network of Excellence.

Accepted for publication in: British Journal of Clinical Pharmacology

EEliflif FatmaFatma SSenen BW.inddBW.indd 7733 003-01-113-01-11 15:1715:17 What is already known · Cough and cold medicines are frequently used in children to treat upper respiratory tract infections without solid proof of benefi ts. · Safety issues have been raised about the use of these drugs in young children · In 2007 international warnings have been issued advising against use of these drugs in young children

What this study adds · Cough and cold medicines prescribing by primary care physicians has not really been infl uenced by international warnings in the Netherlands, where no additional national warnings were made and only partially in Italy. · A concerted action should be taken in Europe to strongly advice against the OTC use and prescription of cough and cold medicines in young children.

EEliflif FatmaFatma SSenen BW.inddBW.indd 7744 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 7755 003-01-113-01-11 15:1715:17 Abstract

Aim: To assess the infl uence of national and international warnings on the prescription rates of cough and cold medicines (CCMs) in the youngest children (<2 years) in the Netherlands (NL) and Italy (IT) countries. Methods: Analysis of outpatient electronic medical records of children <2 years in IT and NL. Age and country specifi c prescription prevalence rates were calculated for the period 2005- 2008.Comparisons of prescription rates in 2005 (pre) and 2008 (post) were done by means of a Chi-square test. Results: The cohort consisted of 99,176 children < 2 years of age. After international warnings, overall prescription rates for CCMs decreased slightly from 83 to 77/1000 person years (p=0.05) in Italy and increased in NL from 74 to 92/1000 children per year. Despite the international warnings, prescription rates for nasal sympathomimetics and opium alkaloids increased in NL (p<0.01). In Italy a signifi cant decrease of the rates of opium alkaloids and other cough suppressants (p<0.01) was observed, and also a signifi cant reduction in use of combinations of nasal sympathomimetics. Conclusion: Despite the international safety warnings and negative benefi t-risk profi les, prescription rates of cough and cold medicines remain substantial and were hardly aff ected by the warnings, especially in the Netherlands where no warning was issued. The hazards of use of these medicines in young children should be explicitly stipulated by European Medicines Agency and all national agencies, in order to increase awareness amongst physicians and caretakers and reduce heterogeneity across the EU.

EEliflif FatmaFatma SSenen BW.inddBW.indd 7766 003-01-113-01-11 15:1715:17 Chapter 2.4 : Eff ects of safety warnings on prescription rates of cough and cold medicines in children

Introduction

Cough and Cold Medicines (CCMs) are frequently used to treat upper respiratory tract symptoms in children. This group of medicines includes expectorants, mucolytics, opium alkaloids (for cough), nasal sympathomimetics and anti-histamines. CCMs carry a risk of serious and potentially life-threatening adverse events, such as cardiac arrhythmias, depressed levels of consciousness, and encephalopathy, and should be avoided in the very young [1]. As a result of these safety issues, in 2007 various regulatory actions have been taken in many countries: in the United States, the Food and Drug Administration (FDA) has issued a recommendation advising parents and caregivers not to use these drugs in children younger than 2 [2]; in the United Kingdom, the Medicine and Healthcare Products Regulatory Agency (MHRA) is advising against the use of many CCMs in children under the age of 6 [3], and Health Canada is requiring manufacturers to re-label certain over-the-counter (OTC) CCMs to indicate that these should not be used in children under the age of 6 [4]. In Italy a warning was issued in September 2007 only regarding the use of nasal sympathomimetics in children <12 years. However in the Netherlands and many other countries, no national warning was issued. The safety concerns on the use of CCMs in young children, and the regulatory actions, also received a lot of attention by the scientifi c community. Sharfstein et al published an important paper in the New England Journal of Medicine which was followed by publications in many lay press journals and media attention in most countries [1]. Most of the utilization data that have been published focus on OTC use of CCMs, as this is the major market in the USA [5-10]. Although CCMs can be bought as OTC in many countries, they are also regularly prescribed by physicians, especially when they would be reimbursed. Previ- ous studies have shown that prescription rates of e.g. antidepressants and antipsychotics drop after regulatory warnings. However, to our knowledge, no studies have been performed that evaluated the eff ect of regulatory warnings on CCMs prescription rates [11-13]. We aimed to investigate the extent of use and to study the eff ects of the warnings on the prescription rates of CCMs to children under the age of two (since this is the group included in all international warnings) in the Netherlands and Italy, two countries with diff erent types of direct national warnings (NL none, and Italy only on nasal sympathomimetics).

EEliflif FatmaFatma SSenen BW.inddBW.indd 7777 003-01-113-01-11 15:1715:17 Methods

Data collection Data were obtained from general practice medical record databases in two countries and were extracted and elaborated according to a common protocol. Databases comprised the Pedianet database in Italy [14] and the Integrated Primary Care Information (IPCI) database in the Neth- erlands [15]. These databases include complete automated medical records of primary care physicians. The patient population in each database is representative of the respective Dutch and Italian population regarding age and gender. The electronic records contain anonymous and coded information on patient demographics, symptoms and diagnoses, referrals, clinical fi ndings, laboratory assessments, drug prescriptions, and hospitalisations. Details of these databases regarding their pediatric data are described elsewhere [16].

Study population The study period was from 1 January 2005 to 31 December 2008. The study population in each country consisted of all children aged 0-2 year (up to 24 months) Each patient was followed from start of the study period or date of registration with the primary care practice (whichever was latest), until the patient left the practice or end of the study period (whichever was earliest).

Classifi cation of prescriptions Cough and cold medicines of interest comprised the following medicines: expectorants (ATC R05CA: althea root, thyme, combinations containing promethazine, oxomemazine, terpenes, guaiacol, and eucalyptus), mucolytics (ATC R05CB: ambroxol, bromhexine, acetylcysteine, carbocisteine, sobrerol, erdosteine, dornase alpha), opium alkaloids and derivatives (ATC R05DA: codeine, dextromethorphan, dihydrocodeine, noscapine, and opium combinations containing pentetrazol, pseudoephedrine, triprolidine, or codeine), other cough suppressants (ATC R05DB: pentoxyverine, levodropropizine, clobutinol, butarimate, dropropizine, cloperastine, and combinations containing pentoxyverine, terpenes, or thyme), nasal sympathomimetics (ATC R01AA: xylometazoline, oxymetazoline, phenylephrine, naphazoline) and combinations of nasal sympathomimetics (ATC R01AB). These drug classes were extracted from the databases using the World Health Organization’s Anatomical Therapeutic Chemical (ATC) classifi cation system [17].

Statistical analysis Annual prescription rates (users per 1000 person years [PY]) of CCMs were estimated by counting the number of children being prescribed each drug or drug class divided by the number of person years attributed, categorized by calendar year, age, and country. Prescription rates can be interpreted as the number of children per 1,000 children who get a specifi c drug (class) prescribed in one year. Because we used annual prescription rates, a child that uses one drug

EEliflif FatmaFatma SSenen BW.inddBW.indd 7788 003-01-113-01-11 15:1715:17 Chapter 2.4 : Eff ects of safety warnings on prescription rates of cough and cold medicines in children

(class) multiple times in one year will be counted only once for that year. To inspect seasonal variability monthly rates were calculated by dividing the number of users per month by the person months attributed. To study the eff ect of the international warnings in 2007 on prescription rates, we compared prescription rates between 2005 and 2008 for each individual drug and drug class by a Chi- square test. P-values below 0.05 were considered to be signifi cant.

Results

Patient characteristics Our population of 99,176 children under the age of 2 generated 110,169 person years of follow-up (Table 1). The overall distribution of children per country was fairly equal, 52% of the patients were from Italy (IT) and 48% from the Netherlands (NL).

Table 1: Number of children in the study population and persontime by gender and calendar year Italy NL Patient Children# PY GP Children# PY GP characteristics practices* practices* Females 24532 (48%) 33862 23406(49%) 19741 Males 26890 (52%) 37002 24348 (51%) 19564

2005 19070 16338 158 11331 8422 134 2006 19424 18780 176 12189 8340 136 2007 18611 18234 160 13128 10845 138 2008 17596 17512 170 13219 11698 138 Total 51422 70864 47754 39305 PY=person-years; *= the number of contributing general practitioner’s databases to the study population #The number of children in various years does not add up to the total since one child can contribute in more than one year during the study period.

Prescription rates of cough and cold medicines In the Netherlands CCMs prescription rates increased signifi cantly between 2005 and 2008 from 74 to 92/1000 PY (p<0.01), but decreased slightly from 83 to 77/1000 PY (p=0.05) in Italy (fi gure 1). In both countries a strong seasonal prescription pattern could be observed, with peaks during winter time (fi gure 2). The seasonal patterns of CCMs prescription rates were similar before and after the international warnings in Italy and the Netherlands. When looking at classes of CCMs, diff erent patterns were observed, both in terms of preferred medicines between countries and changes in annual prescription rates. Expectorants were mostly prescribed in NL, and rates decreased signifi cantly from 10.8 to 4.3/1000 PY (p<0.01) (table 2), mostly because of a decrease in prescription of both thyme syrup (p<0.01) and

EEliflif FatmaFatma SSenen BW.inddBW.indd 7799 003-01-113-01-11 15:1715:17 Figure 1: Prescription rates of cough and cold medicines by database and year with 95% confi dence intervals IPCI= Dutch database Pedianet = Italian database

Figure 2: Prescription rates of cough and cold medicines by database and calendar month IPCI= Dutch database Pedianet = Italian database

combination products (p<0.01) (Table 3). In Italy the only expectorants prescribed were combination products, and prescription rates for these increased slightly.

EEliflif FatmaFatma SSenen BW.inddBW.indd 8800 003-01-113-01-11 15:1715:17 Chapter 2.4 : Eff ects of safety warnings on prescription rates of cough and cold medicines in children

Table 2: Prescription rates of cough and cold medicines per 1000 person years by year and country CCMs Cough relievers Cough suppressants Nasal decongestants Opium alkaloids Other Sympatho- Sympathomimetic Overall Expectorants Mucolytics and mimetics Combinations derivatives

NL IT NL IT NL IT NL IT NL IT NL IT NL IT Year 2005 74 83 10.8 4.1 6.6 54 4.2 7.4 7.7 21 51 0.9 - 7.3 2006 91 82 9.4 5.6 7.1 56 6.4 5.4 7.2 22 67 1.1 - 4.6 2007 93 78 6.8 5.2 8.6 58 10.9 3.6 9.7 17 70 0.3 - 1.4 2008 92 77 4.3 5.0 5.2 58 13.4 3.6 6.2 16 73 0.1 - 0.3 Overall 88* 79 7.1* 5.0 6.8 57 9.3* 5.0* 7.7 19* 66* 0.6* - 3.3* NL: the Netherlands, IT: Italy, - = not available in NL *= signifi cant change between 2005 and 2008 (p<0.05)

Table 3: Pre- and post-warning prescription rates per 1000 person years of individual cough and cold medicines Class of CCM Netherlands Italy 2005 2008 p-value 2005 2008 p-value Combinations* 4.3 1.7 <0.01 Combinations# 4.1 5.0 0.24 Expectorants Thyme syrup 14.8 0.8 <0.01 Althea root 0 0.08 0.40 Bromhexine 4.4 3.2 0.16 Ambroxol 17.6 19.8 0.13 Acetylcysteine 2.0 2.0 0.93 Sobrerol 16.0 20.2 <0.01 Mucolytics Carbocisteine 0.2 0 0.47 Carbocisteine 13.4 14.0 0.65 Dornase alpha 0 .08 0.40 Bromhexine 9.7 8.2 0.14 Noscapine 3.7 14.1 <0.01 Dextromethorphan 3.8 1.8 <0.01 Opium alkaloids and Codeine 0.8 0.4 0.88 Codeine 2.3 0.9 <0.01 Derivatives Combinations## 0.7 0.2 0.01 Pentoxyverine 7.4 6.2 0.30 Levodropropizine 14.8 10.8 <0.01 Other cough Combinations** 0.6 0 0.28 Cloperastine 5.6 5.0 0.40 suppressants Clobutinol 0.2 0 0.42 Nasal Xylometazoline 51 73 <0.01 Phenylephrine 0.8 0.1 <0.01 Sympathomimetics Oxymetazoline 0.1 0.4 0.21 Naphazoline 0.1 0 0.48 Combinations - Ephedrine based 4.6 0.2 <0.01 of nasal Tuaminoheptane 2.9 0.1 <0.01 sympathomimetics based - = not available in NL * combinations containing: promethazine, oxomemazine, guaiacol, ipecacuanha ** combinations containing: pentoxyverine, terpenes, thyme # combinations containing: terpenes, eucalyptus, guaiacol, thyme ## combinations containing: pentetrazol, dihydrocodeine, codeine, dextromethorphan, pseudoephedrine, triprolidine

Mucolytics were ten times more often prescribed in Italy than in NL. In both countries, overall rates did not change before and after the warning, but small changes occurred in prescription

EEliflif FatmaFatma SSenen BW.inddBW.indd 8811 003-01-113-01-11 15:1715:17 of individual products. Especially prescription rates of sobrerol increased in Italy and dornase alpha in NL (tables 2 and 3). Opium alkaloids and derivatives for cough suppression were most frequently prescribed in NL and prescription rates even tripled between 2005 and 2008 from 4.2 to 13.4/1000 PY (p<0.01). This increase was due to a substantial rise in prescription rates of noscapine. In contrast, in Italy prescription rates of opium alkaloids dropped by half from 7.4 in 2005 to 3.6/1000 PY in 2008 (p<0.01). This decrease was seen for all products in this class (Tables 2 and 3). Prescription rates for other (non-opioid) cough suppressants were highest in Italy, and this was mostly accounted for by levodropropizine. Rates decreased by one third after the warning (p<0.01 (Table 2 and 3)). In NL, use of other cough suppressants did not change signifi cantly. Single product nasal sympathomimetics (mostly xylometazoline) were primarily prescribed in NL, and prescription rates increased signifi cantly from 51 in 2005 to 73/1000 PY in 2008, mostly due to use of xylometazoline (p<0.01) (Table 2). In IT, xylometazoline was hardly prescribed even before the warning, and prescription rates decreased to almost 0 in 2008. On the other hand, combinations of nasal sympathomimetics were only prescribed in IT, and a strong eff ect of the warning was seen on their prescription rates (from 7.3 in 2005 to 0.3/1000 PY in 2008, p<0.01). Medicines prescribed in this class were combinations with ephedrine or tuaminoheptane.

Discussion

The formal regulatory review of OTC drugs, including CCMs, started in 1972 in the United States [18]. At that time, most OTC CCMs were generally recognized as safe and eff ective. Until recently CCMs received little attention regarding their safety and effi cacy in children. In 2007, this changed dramatically by a citizen petition submitted to the FDA, raising signifi cant concerns about the safety and effi cacy of CCMs in children under the age of 6 years. This petition resulted in a public healt advisory by the FDA, recommending that CCMs should not be used in children under 2 years of age because of the risk of serious and life-threatening eff ects [19]. The FDA warning was followed by other countries issuing warnings and several literature publications. This paper provides information on the eff ects of national and international warnings on rates of prescriptions of contra-indicated drugs in a country with and a country without additional national warnings. Despite the international warnings and important publications in high impact journals and lay press advising against the use of CCMs in young children, overall prescription rates for CCMs increased in the Netherlands, where no national warning was issued. Use of opium alkaloids and nasal sympathomimetics increased signifi cantly, although these drugs are contraindicated in young children [20]. In Italy, where a specifi c warning was issued by the Agenzia Italiana del Farmaco only against the use of nasal sy mpathomimetics, a signifi cant reduction in use of nasal sympathomimetics was seen and also a decrease in use of cough suppressants (opioids and non-opioids).

EEliflif FatmaFatma SSenen BW.inddBW.indd 8822 003-01-113-01-11 15:1715:17 Chapter 2.4 : Eff ects of safety warnings on prescription rates of cough and cold medicines in children

Although the warnings did not show a considerable impact on the overall prescription rates in the Netherlands, there were shifts regarding use of individual medicines. Rates for expectorant combinations decreased (comprising of promethazine and oxomemazine), whereas the prescription rates for noscapine, an opium alkaloid, increased substantially. Substantial evidence of adverse events related to the use of promethazine and oxomemazine in children exists [1, 21]. There are no studies evaluating the safety of noscapine in children, and no safety issues have been raised so far on the use of these drugs in young children. The Dutch GP society recommends noscapine for the symptomatic treatment of cough in young children [22]. This change from expectorant combinations to noscapine may be seen as a favourable trend although safety of noscapine in these young children should be further investigated. The observed increase in prescription rates of nasal sympathomimetics in the Netherlands is of high concern, since these drugs may cause serious adverse events such as lethargy, tachycardia, convulsions, and even coma [23-27]. Furthermore, a recent Cochrane review concluded that insuffi cient data are available on safety and effi cacy of many nasal decongestants in children, which are therefore not recommended for use in children <12 years of age [28].In Italy, safety warnings were issued in September 2007 advising not to use nasal sympathomimetics in children <12 years. Prescription rates of these drugs reduced signifi cantly after the warning. Our data show that CCMs are still prescribed to very young children, and prescription rates were hardly infl uenced by the international safety warnings, especially in the Netherlands where no national regulatory action was taken. In Italy we could observe positive eff ects from the safety warnings in some drug classes. Overall however, more action should be taken since use of these drugs in the very young may raise safety issues, and there is no evidence of effi cacy [21, 29-31]. The few studies that evaluated the effi cacy of CCMs in children with cough and cold found no signifi cant benefi t in symptomatic relief when compared to placebo [32-36]. Beside a lack of effi cacy, CCMs are more prone to dosing errors. Variations in liquid CCMs dosing with spoons can be as much as 20%, leading to a higher risk of adverse events [37]. A recent study investigating therapeutic errors in children found that 23% of the dosing errors were related to use of CCMs [38]. Another study exploring pediatric fatalities associated with OTC CCMs use concluded that 85% of the fatalities were associated with an overdose of CCMs, mainly in children under the age of 2 [39]. Not only the country specifi c diff erences of the international warnings on the prescription rates of CCMs are of interest but also the diff erences in the types of CCMs that are prescribed. In Italy sympathomimetics are hardly used but very commonly prescribed in the Netherlands, whereas in the Netherlands use of mucolytics is very low, but frequently prescribed in Italy. This is probably due to country specifi c guidelines on the symptomatic relief of cough and cold. Many of the CCMs, which are still prescribed as seen in this study, are contraindicated in children younger than 6 years in the United Kingdom and Canada, and younger than 2 in the US [2, 3, 40]. This makes prescription by primary care physicians to young children even more alarming. To minimize the use of CCMs in children, these drugs should no longer be available as OTC; and

EEliflif FatmaFatma SSenen BW.inddBW.indd 8833 003-01-113-01-11 15:1715:17 caregiver and physician education about the self-limiting nature of coughs and colds, the lack of effi cacy and the risk associated with these drugs is needed [1, 41]. As for all observational research, our data has limitations. First of all, since we used primary care prescription data, we did not capture OTC use of these drugs, nor had data on actual intake. However, our aim was to analyse prescription rates of CCMs by primary care physicians as these health professionals are approached directly by regulators/inspectorates. Furthermore, we could not compare our results with pre- and post warning prescription rates in countries where more extensive warnings have been given, since there is no published data about this subject. A major strength of our study is the fact that we capture a large number of children in two diff erent databases across Europe, resulting in a high generalisability of the study population. To conclude, our results show that despite international warnings, a negative safety profi le and lack of effi cacy data, CCMs are still prescribed to young children in primary care in Italy and the Netherlands, and in the latter country rates even increased. Also, CCMs were prescribed that contain specifi c active ingredients proven to be most harmful in young children. Two lessons can be learned from this study. First of all, primary care physicians should be more aware of the harm and lack of effi cacy of these drugs and should not prescribe them to young children. Secondly, there is heterogeneity in warnings across countries in the EU, which creates inequality. In order to increase awareness among physicians and caretakers and to allow for consistent warnings, the hazards of use of these medicines have to be explicitly stipulated by European Medicines Agency. A concerted action is needed in Europe to advice against the use and prescription of cough and cold medicines to young children.

EEliflif FatmaFatma SSenen BW.inddBW.indd 8844 003-01-113-01-11 15:1715:17 Chapter 2.4 : Eff ects of safety warnings on prescription rates of cough and cold medicines in children

References

1 Sharfstein JM, North M, Serwint JR. Over the counter but no longer under the radar--pediatric cough and cold medications. The New England journal of medicine. 2007 Dec 6;357(23):2321-4. 2 FDA. FDA Recommends that Over-the-Counter (OTC) Cough and Cold Products not be used for Infants and Children under 2 Years of Age. 2008 [cited 03-10-2009]; Available from: http://www.fda. gov/Drugs/DrugSafety/PublicHealthAdvisories/ucm051137.html 3 MHRA. Press release: Better medicines for children’s coughs and colds. 2009 [cited 03-10-2009]; Avail- able from: http://www.mhra.gov.uk/NewsCentre/Pressreleases/CON038902 4 Canada H. Health Canada Releases Decision on the Labelling of Cough and Cold Products for Children. 2008 [cited 03-10-2009]; Available from: http://www.hc-sc.gc.ca/ahc-asc/media/advisories- avis/_2008/2008_184-eng.php 5 Children’s OTC cold medicines: the public, and parents, weigh in. 2007 [cited 05-10-2009]; Available from: http://www.kff .org/kaiserpolls/upload/7725.pdf 6 Gadomski AM, Rubin JD. Cough and cold medicine use in young children: a survey of Maryland pediatricians. Md Med J. 1993 Jul;42(7):647-50. 7 Headley J, Northstone K. Medication administered to children from 0 to 7.5 years in the Avon Lon- gitudinal Study of Parents and Children (ALSPAC). European journal of clinical pharmacology. 2007 Feb;63(2):189-95. 8 Kogan MD, Pappas G, Yu SM, Kotelchuck M. Over-the-counter medication use among US preschool- age children. Jama. 1994 Oct 5;272(13):1025-30. 9 Vernacchio L, Kelly JP, Kaufman DW, Mitchell AA. Pseudoephedrine use among US children, 1999- 2006: results from the Slone survey. Pediatrics. 2008 Dec;122(6):1299-304. 10 Vernacchio L, Kelly JP, Kaufman DW, Mitchell AA. Cough and cold medication use by US children, 1999-2006: results from the slone survey. Pediatrics. 2008 Aug;122(2):e323-9. 11 Katz LY, Kozyrskyj AL, Prior HJ, Enns MW, Cox BJ, Sareen J. Eff ect of regulatory warnings on antidepressant prescription rates, use of health services and outcomes among children, adolescents and young adults. Cmaj. 2008 Apr 8;178(8):1005-11. 12 Kurian BT, Ray WA, Arbogast PG, Fuchs DC, Dudley JA, Cooper WO. Eff ect of regulatory warnings on antidepressant prescribing for children and adolescents. Archives of pediatrics & adolescent medicine. 2007 Jul;161(7):690-6. 13 Valiyeva E, Herrmann N, Rochon PA, Gill SS, Anderson GM. Eff ect of regulatory warnings on antipsychotic prescription rates among elderly patients with dementia: a population-based time-series analysis. Cmaj. 2008 Aug 26;179(5):438-46. 14 Menniti-Ippolito G, Raschetti R, Da Cas R, Giaquinto C, Cantarutti L. Active monitoring of adverse drug reactions in children. Italian PaediatricPharmacosurveillance Multicenter Group. Lancet. 2000;355:1613-4. 15 Vlug AE, van der Lei J, Mosseveld BM, van Wijk MA, van der Linden PD, Sturkenboom MC, et al. Post- marketing surveillance based on electronic patient records: the IPCI project. Methods of information in medicine. 1999 Dec;38(4-5):339-44. 16 Sturkenboom MC, Verhamme KM, Nicolosi A, Murray ML, Neubert A, Caudri D, et al. Drug use in children: cohort study in three European countries. BMJ (Clinical research ed. 2008;337:a2245. 17 WHO Collaborating Center for Drug Satistics Methodology. [cited; Available from: http://www.whocc. no/atcddd/ 18 Vassilev ZP, Kabadi S, Villa R. Safety and effi cacy of over- the-counter cough and cold medicines for use in children. Expert opinion on drug safety. Mar;9(2):233-42.

EEliflif FatmaFatma SSenen BW.inddBW.indd 8855 003-01-113-01-11 15:1715:17 19 FDA. FDA Recommends that Over-the-Counter (OTC) Cough and Cold Products not be used for Infants and Children under 2 Years of Age. http://wwwfdagov/Drugs/DrugSafety/PublicHealthAdvisories/ ucm051137html 2008 [cited 03-10-2009]; http://www.fda.gov/Drugs/DrugSafety/PublicHealthAdvi- sories/ucm051137.html 20 Kuehn BM. Citing serious risks, FDA recommends no cold and cough medicines for infants. Jama. 2008 Feb 27;299(8):887-8. 21 Rimsza ME, Newberry S. Unexpected infant deaths associated with use of cough and cold medications. Pediatrics. 2008 Aug;122(2):e318-22. 22 Canada H. Health Canada Advisory: Cough and Cold products for children. http://chdregionwa- terlooonca/web/healthnsf/vwSiteMap/E7742A6975D6895E852570E50055EEFF/$file/PHYSUP_ JAN09pdf?openelement 2009 [cited 07-10-2009]; http://chd.region.waterloo.on.ca/web/health.nsf/ vwSiteMap/E7742A6975D6895E852570E50055EEFF/$fi le/PHYSUP_JAN09.pdf?openelement 23 Dunn C, Gauthier M, Gaudreault P. Coma in a neonate following single intranasal dose of xylometazoline. European journal of pediatrics. 1993 Jun;152(6):541. 24 Higgins GL, 3rd, Campbell B, Wallace K, Talbot S. Pediatric poisoning from over-the-counter imidazoline-containing products. Annals of emergency medicine. 1991 Jun;20(6):655-8. 25 Mahieu LM, Rooman RP, Goossens E. Imidazoline intoxication in children. European journal of pediatrics. 1993 Nov;152(11):944-6. 26 Soderman P, Sahlberg D, Wiholm BE. CNS reactions to nose drops in small children. Lancet. 1984 Mar 10;1(8376):573. 27 van Velzen AG, van Riel AJ, Hunault C, van Riemsdijk TE, de Vries I, Meulenbelt J. A case series of xylometazoline overdose in children. Clinical toxicology (Philadelphia, Pa. 2007;45(3):290-4. 28 Taverner D, Latte GJ. WITHDRAWN: Nasal decongestants for the common cold. Cochrane database of systematic reviews (Online). 2009(2):CD001953. 29 Infant deaths associated with cough and cold medications--two states, 2005. Mmwr. 2007 Jan 12;56(1):1-4. 30 Wingert WE, Mundy LA, Collins GL, Chmara ES. Possible role of pseudoephedrine and other over-the- counter cold medications in the deaths of very young children. Journal of forensic sciences. 2007 Mar;52(2):487-90. 31 Marinetti L, Lehman L, Casto B, Harshbarger K, Kubiczek P, Davis J. Over-the-counter cold medications-postmortem fi ndings in infants and the relationship to cause of death. Journal of analytical toxicology. 2005 Oct;29(7):738-43. 32 Smith SM, Schroeder K, Fahey T. Over-the-counter medications for acute cough in children and adults in ambulatory settings. Cochrane database of systematic reviews (Online). 2008(1):CD001831. 33 Smith MB, Feldman W. Over-the-counter cold medications. A critical review of clinical trials between 1950 and 1991. Jama. 1993 May 5;269(17):2258-63. 34 Taylor JA, Novack AH, Almquist JR, Rogers JE. Effi cacy of cough suppressants in children. The Journal of pediatrics. 1993 May;122(5 Pt 1):799-802. 35 Paul IM, Yoder KE, Crowell KR, Shaff er ML, McMillan HS, Carlson LC, et al. Eff ect of dextromethorphan, diphenhydramine, and placebo on nocturnal cough and sleep quality for coughing children and their parents. Pediatrics. 2004 Jul;114(1):e85-90. 36 Paul IM, Beiler J, McMonagle A, Shaff er ML, Duda L, Berlin CM, Jr. Eff ect of honey, dextromethorphan, and no treatment on nocturnal cough and sleep quality for coughing children and their parents. Archives of pediatrics & adolescent medicine. 2007 Dec;161(12):1140-6. 37 Wansink B, van Ittersum K. Spoons systematically bias dosing of liquid medicine. Annals of internal medicine. Jan 5;152(1):66-7.

EEliflif FatmaFatma SSenen BW.inddBW.indd 8866 003-01-113-01-11 15:1715:17 Chapter 2.4 : Eff ects of safety warnings on prescription rates of cough and cold medicines in children

38 Taylor DM, Robinson J, MacLeod D, MacBean CE, Braitberg G. Therapeutic errors involving adults in the community setting: nature, causes and outcomes. Australian and New Zealand journal of public health. 2009 Aug;33(4):388-94. 39 Dart RC, Paul IM, Bond GR, Winston DC, Manoguerra AS, Palmer RB, et al. Pediatric fatalities associated with over the counter (nonprescription) cough and cold medications. Annals of emergency medicine. 2009 Apr;53(4):411-7. 40 Canada H. Health Canada Advisory: Cough and Cold products for children. 2009 [cited 07-10-2009]; Available from: http://chd.region.waterloo.on.ca/web/health.nsf/vwSiteMap/E7742A6975D6895E85 2570E50055EEFF/$fi le/PHYSUP_JAN09.pdf?openelement 41 Use of codeine- and dextromethorphan-containing cough remedies in children. American Academy of Pediatrics. Committee on Drugs. Pediatrics. 1997 Jun;99(6):918-20.

Acknowledgements: We thank all of the physicians contributing data to the PEDIANET, and IPCI databases.

Funding: Funded by the European Community’s 6th Framework Programme. Project number LSHB-CT-2005-005216: TEDDY: Task force in Europe for Drug Development for the Young.

EEliflif FatmaFatma SSenen BW.inddBW.indd 8877 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 8888 003-01-113-01-11 15:1715:17 Chapter 3

Drug safety

EEliflif FatmaFatma SSenen BW.inddBW.indd 8899 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 9900 003-01-113-01-11 15:1715:17 Chapter 3.1

Use of Asthma medicines and reported adverse drug reactions in children: Results from the WHO adverse events database

E. Fatma Sen, Katia MC Verhamme, Sandra de Bie, Bruno HCh Stricker, Carlo Giaquinto, Adriana Ceci, and Miriam CJM Sturkenboom, on behalf of the TEDDY European Network of Excellence.

EEliflif FatmaFatma SSenen BW.inddBW.indd 9911 003-01-113-01-11 15:1715:17 Abstract

Objective: To provide an overview of the paediatric adverse drug reactions (ADR), related to the use of asthma drugs in children, that have been reported to the WHO adverse events reporting database. In addition, we want to identify new ADRs associated with use of asthma medicines in children. Design: case non-case study Setting: worldwide spontaneous reporting database: Vigibase Participants: children aged 0-≤18 years Main outcome measure: association between asthma medicines and suspected ADRs using the reporting odds ratio (ROR) as a measure of disproportionality while adjusting for gender, continent and reporter. To reduce confounding , vaccine-related records were excluded from the analysis. Results: Records for asthma drug related ADRs comprised 3.5% of the non-vaccine related records (10,259 records). The majority of the records were for children aged 2-≤11 years (60%) and most records concerned short-acting β2-agonists (SABAs) and inhaled corticosteroids. Known ADRs such as Churg -Strauss syndrome, paradoxical bronchospasm and adrenal insuffi ciency were amongst the ADRs with the highest RORs. The majority of the identifi ed asthma medicines-ADR associations are already described in the literature. New, potential signals that we found were intestinal gangrene (ROR 38.2, 95% CI [17-267]) and perforated gastric ulcer (ROR 55.3, 95% CI [21-148]) both associated with use of SABAs, and tooth discoloration with the use of anticholinergics (ROR 21.4, 95% CI [8-59]). Conclusion: this systematic assessment of pediatric ADRs for asthma medicines in children shows that the majority of ADRs reported to Vigibase are already described in the literature. Some potentially new ADRs were identifi ed, which require further exploration.

EEliflif FatmaFatma SSenen BW.inddBW.indd 9922 003-01-113-01-11 15:1715:17 Chapter 3.1: Asthma medicines and reported ADRs in children

Introduction Respiratory medicines are amongst the most frequently prescribed drugs in children, and especially asthma medicines are frequently prescribed [1]. Prevalence rates of asthma medicines use range between 4 and 26 percent, depending on country, age range, and study period [1-3]. According to the Global Initiative for Asthma (GINA), bronchodilators, (inhaled short-acting ß2-agonists, long-acting ß2-agonists, anticholinergics, inhaled corticosteroids and alternative (add-on) treatments such as leukotriene receptor antagonists (LTRA), xanthines and anti- allergics (cromones) all have a place in the treatment of childhood asthma, although at various stages depending on severity of illness [4]. As asthma is the most common chronic disease of childhood, most of these medicines is used for a long term and for some of them, safety concerns exist with long duration of use [5]. For example, increased risk of death with single use of long-acting beta-agonists in adults, and adrenal suppression and growth retardation with use of inhaled steroids in children [6, 7]. However, overall the risk-benefi t profi le of these medicines has been well studied in randomized controlled trials (RCTs), and is considered to be favorable [4, 7]. However, the majority of these studies have been performed in children older than 5 years and studies in younger children are lacking. Furthermore, RCTs have small sample sizes and relatively short follow-up, which makes it diffi cult to detect adverse drug reactions (ADRs) which are rare or those that occur after long-term treatment. The missing data on safety and effi cacy information regarding these medicines has been a reason for the European Medicines Agency (EMA) to put many asthma medicines on the pediatric needs list. This list, drafted by Pediatric Committee of the EMA, contains medicines for which information on pharmacokinetics, effi cacy and safety is urgently needed [8]. Large spontaneous adverse event reporting databases, such as the Vigibase database of the WHO-UMC (Uppsala Monitoring Centre), is a well-recognized source for monitoring the potential risks of drugs, however there has been no systematic approach to investigate safety signals in children. To get a broader understanding of the adverse drug reactions associated with asthma medicines use, which is one of the most frequently used drug group in children, we analysed the adverse drug reactions in relation to asthma medicines reported to the Vigibase database of the WHO-UMC [1].

EEliflif FatmaFatma SSenen BW.inddBW.indd 9933 003-01-113-01-11 15:1715:17 Methods

WHO-UMC database We used data from the Vigibase database of suspected ADRs. This global individual case safety report (ICSR) database system, established by the World Health Organisation (WHO) in 1968, holds more than 3.7 million ICSRs as of June 2006. It is maintained on behalf of the WHO Programme by the Uppsala Monitoring Centre (UMC). More than 80 countries participate in the WHO International Drug Monitoring program, and another 17 countries are associate members, not yet actively contributing data. ICSRs are submitted through the national pharmacovigilance centers. WHO Programme member countries submit ICSRs to the UMC on a regular basis; preferably once per month, but at least every quarter.[9].

At the time of the data extraction in 2006, all ADRs within Vigibase were coded using preferred terms of the WHO-Adverse Reaction Terminology (WHO-ART) coding dictionary. Reported drugs were re-coded using the WHO Drug dictionary, and the Anatomical Therapeutic Chemi- cal (ATC) classifi cation system of the WHO Collaborating Centre for Drug Statistics Methodol- ogy [10]. Assigned ATC codes of asthma medicines were verifi ed for accuracy in this study. For the purpose of this study drugs were further categorized as: short-acting β2-agonists (SABA, ATC codes R03AC02, R03AC03 or R03AC04), long-acting β2-agonists (LABA, ATC codes R03AC excluding SABA), inhaled glucocorticoids (ICS, ATC R03BA), anticholinergics (ACH, ATC R03BB), anti-allergic agents (anti-allergics, ATC R03BC), xanthines (ATC R03DA), leukotriene receptor antagonists (LTRA, ATC R03DC), fi xed combinations of β2-agonists+ICS (β2+ICS, ATC R03AK06 or R03AK07) and fi xed combinations of β2-agonists+ACH (β2+ACH, R03AK03 or R03AK04).

Selection of ICSRs From Vigibase, all ICSRs in children (age 0 - ≤18 years) with onset date between January 2000 and December 2006 were extracted. Only spontaneously reported ICSRs and those with a suspected causal relationship between ADR and drug were included. ICSRs for which information on the reported drug or reported ADR was missing were excluded.

Record defi nition One ICSR can contain more than one suspected drug, and more than one ADR. We defi ned a record as a unique combination of a drug and an ADR. Hence, an ICSR containing two ADRs with one suspected drug counted for two records, and an ICSR containing two ADRs with two suspected drugs counted for four records. (In Sandra’s paper kreeg ze vragen over deze keuze van selectie van data – kan je evtl verwijzen naar paper van WHO waar ze ook deze method aanhouden?) Information on each record include country of origin, reporter, age at onset, year of onset, gender, reported drug, reported ADR, starting and stopping date of the drug, starting and stopping date of the ADR, dosing regimen of the drug, administration route, and causality assessment.

EEliflif FatmaFatma SSenen BW.inddBW.indd 9944 003-01-113-01-11 15:1715:17 Chapter 3.1: Asthma medicines and reported ADRs in children

Statistical analysis All analyses were conducted on a record basis, thus using unique drug-ADR combinations. Disproportionality was assessed by calculation of the reporting odds ratio (ROR), which is analogous to an odds ratio [11]. RORs with 95% confi dence intervals (95% CI) were calculated for the diff erent subgroups of cough and cold medicines. The ROR was only calculated if at least 3 records of the drug-ADR combination of interest were available. A ROR of ≥ 2 with a 95% confi dence interval not including 1 is considered a signal . Since this method is purely hypothesis generating it must be emphasized that a signifi cant signal does not imply causality. To study the eff ect of age; the ROR was calculated within predefi ned age-categories. In line with the guidelines of the International Conference of Harmonization (ICH), age at the time of the event was categorized into the following age categories: 0 - < 2 years, 2 - ≤11 years and 12 - ≤18 years [12]. Since vaccines represent a substantial part of the drugs used in children and may confound the RORs of other drug-ADR combinations, vaccine related records were excluded. Additional logistic regression analyses using SPSS version 15.0 for Windows (SPSS, Inc., Chicago, Illinois) was performed to correct for gender, continent and type of reporter.

Results

In the period between January 2000 and December 2006 221,508 suspected (thus potentially causally related) spontaneous ICSRs in children aged ≤18 years were included in the WHO- Vigibase translating in a total of 845,090 records. After exclusion of records that were vaccine related (n= 517,642) and for which information on the reported drug or reported ADR was missing (n= 32,675), the total number of drug-ADR combinations in the analysis was 294,773. The database comprised 10,259 asthma medicines related records (3.5% of all non-vaccine related records). Compared to all non-vaccine related records, records of asthma medicines related ADRs involved more boys, were more often reported by a general practitioner, and involved younger children (mean age diff erence of 1.3 years) (table 1). ADRs were most frequently reported for SABA (24.1%) ICS (23.6%) and LTRAs (20.8%) (table 2).

Drug-ADR combinations with the highest RORs Table 3 shows the top ten of strongest associations. The number one was Churg-Strauss syndrome (ROR 365, 95% CI 48-2802) followed by paradoxical bronchospasm (ROR (77 95% CI 21-284) (Table 3) and cortisol decreased. Most of the associations were not seen in the youngest children, apart from decreased levels of cortisol and tenosynovitis.

EEliflif FatmaFatma SSenen BW.inddBW.indd 9955 003-01-113-01-11 15:1715:17 Table 1: Characteristics of all non-vaccine and asthma medicines related ADR records % of total Asthma % of all All non-vaccine related number of medicines records records records related records (row) N 294,773 10,259

Male gender 153,708 52.1 5,787 3.8

Notifi er General practitioner 53,790 18.2 2,438 4.5 Physician 50,300 17.1 1,368 2.7 Hospital 38,795 13.2 349 0.9 Other 151,888 51.5 5,834 3.8

Country of origin United States 181,126 61.4 6,367 3.5 France 18,820 6.4 374 2.0 Thailand 16,464 5.6 182 1.1 Other//

Age Mean / SD (years) 9.7 6.0 8.4 Median (years) 10.0 8.0 0 - < 2 years 45,209 15.3 1,023 2.3 2 - ≤ 11 years 114,704 38.9 6,199 5.4 12 - ≤ 18 years 134,860 45.8 3,037 2.3

Table 2: Distribution of asthma drugs related records by drug class and age category Drug class 0-2 years 2-11 years 11-18 years Total N %* N %* N %* N%* SABA 427 41.7 1,235 19.9 806 26.5 2,468 24.1 LABA 6 0.6 270 4.4 302 9.9 578 5.6 ICS 238 23.3 1,698 27.4 480 15.8 2,416 23.6 ACH 30 2.9 90 1.5 10 0.3 130 1.3 Anti-allerigcs 27 2.6 203 3.3 54 1.8 284 2.8 Xanthines 110 10.8 211 3.4 154 5.1 475 4.6 LTRA 76 7.4 1,505 24.3 557 18.3 2,138 20.8 β-agonists+ICS 10 1.0 747 12.1 440 14.5 1,197 11.7 β-agonists+ACH 12 1.2 17 0.3 8 0.3 37 0.4 Other asthma drugs 87 8.5 223 3.6 226 7.4 536 5.2 Total 1,023 100 6,199 100 3,037 100 10,259 100 * = % of total within this age category SABA=short-acting beta-agonists, LABA=long-acting beta-agonists, ICS = inhaled corticosteroids, ACH=anticholinergics, LTRA = leukotriene receptor antagonists

Analyses by type of asthma medicines showed that for SABAs paradoxical bronchospasm, intestinal gangrene and perforated gastric ulcer were associated with the highest RORs (table 4). In

EEliflif FatmaFatma SSenen BW.inddBW.indd 9966 003-01-113-01-11 15:1715:17 Chapter 3.1: Asthma medicines and reported ADRs in children

Table 3: Ten highest reporting odds ratios (RORs) for ADRs associated with asthma drugs stratifi ed by age All age categories Stratifi ed by age category Associated Associated with ROR* ROR* ROR* Adverse drug with any ROR crude ROR* asthma (95% CI) (95% CI) (95% CI) reaction other drug (95% CI) (95% CI) drugs 0-2 yr 2-≤11 yr 12-≤18 yr N N Churg Strauss 352.6 365.8 244.1 13 1 - Infi nity syndrome (46-2696) (48-2802) (29-2031) Bronchospasm 90.4 77.7 23.3 10 3 - Infi nity paradoxical (25-328) (21-284) (5-99) 81.5 73.8 71.8 105.8 Cortisol decreased 30 10 - (40-167) (36-151) (21-251) (41-271) Adrenal 67.5 63.1 19.9 96.7 84 34 17.5 (7-46) Insuffi cieny (45-101) (42-94) (6-62) (52-79) Intestinal 28.4 49.1 5 4 33.9 (9-126) -- gangrene (8-107) (11-210) Nasal septum 32.7 7 6 31.6 (11-94) 24.1 (8-72) - 30.4 (8-114) perforation (3-323) 29.0 Anisocoria 4 4 27.1 (7-108) 23.3 (6-94) - - (5-161) 24.9 (6- Tenosynovitis 4 4 27.1 (7-108) Infi nity -- 100) Bronchospasm 25.4 (13- 24.6 16 18 24.1 (12-47) - 23.2 (9-62) aggrevated 50) (8-76) 20.6 (13- 30.3 Osteoporosis 36 45 21.7 (14-34) - 21.7 (12-39) 32) (14-65) * = adjusted for gender, continent and type of reporter - = number drug-ADR combination with ≥ 3 records

age-specifi c analyses the highest RORs were for psychosis in children aged 0-2 years, bronchiectasis in children 2-≤11 years and intestinal gangrene in the oldest age category. No LABA related ADRs were reported for children aged 0-2 years, which is not indicated below age 5. In children aged 2-≤11 years the highest RORs were for tooth disorders and hemiplegia. In children aged 12-≤18 years, sudden death was amongst the ADRs with the third highest ROR within this drug group (table 4) In all 3 three age categories, disorders of the adrenal gland (adrenal insuffi ciency and decreased levels of cortisol) were ADRs with the highest ROR for ICS. Perforation of the nasal septum was also among the ADRs with the highest ROR in this drug group in children aged 2-≤11 and 12-≤18 years. (table 4) Other ADRs with the highest ROR for ICS were hypertrichosis and varicella in children <2 years and osteoporosis in children aged 12-≤18 years. For anticholinergics respiratory acidosis was strongly associated. When stratifying by age categories, drug-ADR combinations with ≥ 3 records for anticholinergics were only observed in children aged 2-≤11. The three highest RORs in this category were found for tooth discoloura- tion, mydriasis and constipation. (table 4)

EEliflif FatmaFatma SSenen BW.inddBW.indd 9977 003-01-113-01-11 15:1715:17 or SPC literature Conf./Tr. fail. Conf./Tr. Described in Conf. /Tr. fail. /Tr. Conf. ROR* # no - cases # 4 461172No 76 No 4 82 204 [24] 5 171 [25] 14 8 395 [15] cases ADR Taste Taste perversion Cortisol decreased Nasal septum perforation Intestinal gangrene Perforated gastric ulcer [35] Conf. or SPC literature Described in ROR* # no cases # 4 193 18 16 15 [36] No Asthma 523 214 14 40 292 109 18 No No Osteoporosis 17 31 133 [15] 14 369 258 16 252 [15] [15] cases ADR ciency Tooth disorderTooth 3Hemiplegia 35 3 30 37 No 31 Asthma No 14 252 24 Myocardial Myocardial ischaemia Heart valve disorders AsthmaCortisol decreased Adrenal 11insuffi 336Nasal septum perforation 12 Tooth fail. /Tr. Conf. Sudden deathdiscolouration 3 73 16 [37] MydriasisConstipation 4 3 303 274 15 11 [24] [24] in or SPC literature Described ROR* # no cases 0-2 years 2-11 years 12-18 years # 5 8 59 [15] cases drug class and age category RORs by cant ADR ciency Psychosis 3 2 139 [34] Bronchiectasis 3 13 19 Vasodilatation 3Hypokalemia 14 4 22 [36] 23 19 [36] Hypertrichosis 5Varicella 9 80 3 [38] 13 27 [39] Adrenal Adrenal insuffi Drug class SABA LABA - ICS ACH - Table 4: Three highest signifi Three 4: Table

EEliflif FatmaFatma SSenen BW.inddBW.indd 9988 003-01-113-01-11 15:1715:17 Chapter 3.1: Asthma medicines and reported ADRs in children [28] [29] [29] Conf. - 5 2 654 [30] 53470No ukotriene receptor-antagonists Churg strauss Churg syndrome Glucocorticoids increased nity [30] 3 113 14 No Paraesthesia 3 5753 11 10 [40] 21 other drugs Vasculitis 4 77 8.3 [30] 7-Infi 31818 Photosensitivity reaction Myalgia 3 196 8 [40]Airway obstruction Dyspnoea 8 1,732 12 fail. Conf./Tr. Churg strauss Churg syndrome Medical device complication EnanthemaAsthma 4 29 32 315 21 13 [42] fail. /Tr. Conf. Asthma Moniliasis 22 8 244 69 24 fail. /Tr. Conf. 34 [42] Coughing 4 625 57 [33] in adults Hepatitis er - - 3 117 16 3 1813 10 596 [30] 3 Monocytosis [41] 3 9 19 No Iritis 4 6 137 No 10 32 18 No Hypokinesia 4 108 22 No Acidosis 9 231 24 ParaesthesiaTachycardia 3 8 23 292 50 11 [29] Palpitation Acidosis 3 5 157 186 10 15 [29] [28] Extrasystoles Hypokalaemia 3 5 105 156 25 25 Jaundice Jaundice neonatal Drug level Drug level decreased Allergic reaction Premature birth Pneumonia 3 1054 6 [27] Rhinitis 8 241 17 Conf. Rhinitis 3 234 26 LTRA Xanthines β-agonists+ ICS Antiallergics β-agonists+ ACH * = adjusted for gender, continent and type continent of notifi gender, for * = adjusted - = number of drug-ADR combination with ≥ 3 records with ≥ 3 records - = number of drug-ADR combination = le LTRA SABA=short-acting ACH=anticholinergics, LABA=long-acting ICS = inhaled glucocorticoids, beta-agonists, beta-agonists, Table 4: Continue Table failure = confounding/treatment fail. Conf./Tr.

EEliflif FatmaFatma SSenen BW.inddBW.indd 9999 003-01-113-01-11 15:1715:17 Antiallergics were strongly associated with rhinitis in children aged 2-≤11 and 12-≤18 years. In children <2 years, the only ADR with ≥3 records for antiallergics was pneumonia. For xanthines, heart rhythm disorders related ADRs (tachycardia, palpitation, extrasystoles) were amongst the ADRs with highest RORs in all age categories. (table 4) Acidosis had the highest ROR in children aged 2-≤11 and 12-≤18 years. (table 4) In children < 2 years, the highest RORs for LTRAs were calculated for neonatal jaundice, allergic reactions and premature birth. Churg Strauss syndrome had the highest ROR in children aged 2-≤11 and 12-≤18 years. In children aged 2-≤11 years LTRAs were the only group for which Churg Strauss syndrome was reported as an ADR. In the overall analysis, leg cramps were most strongly associated with β2+ICS combinations. Within the age categories, drug-ADR combinations with ≥ 3 records for β2+ICS were only observed in children aged 2-≤11 and 12-≤18 years. The highest ROR in children 2-≤11 years were observed for medical device complications and for increased levels of glucocorticoids in children aged 12-≤18 years. Flushing and coughing were the ADRs with the highest RORs for β2+ACH in the overall analysis. Within the age categories, the only drug-ADR combination with ≥ 3 records was coughing in children aged 2-≤11 years.

Discussion

This is the fi rst study that analysed pediatric ADRs reported for asthma medicines in the Vigi- base WHO-UMC database. There are several important fi ndings in our study. First, asthma drugs related ADRs were infrequently reported (3.5% of the non-vaccine related records) although drug utilization studies have shown that 4-26% of children use an asthma medicine per year [1, 2]. The low reporting rate could be due to the fact that most asthma drugs are on the market for many years, and are thought to have a good safety profi le. Second, most asthma medicines related ADRs were reported for children aged 2-≤11 and 12-18 years of age although utilization studies have shown that prevalence of use of asthma medicines is highest in younger children [1, 3]. There thus seems to be a discrepancy between the percentage of asthma medicines use in children <2 and the proportion of ADRs reported for this age group. A possible explanation could be that ADRs are not recognized in very young children. Third, most asthma medicines related ADRs concerned SABAs and ICS, drugs for which utilization studies have shown that these are the mostly used asthma medicines in pediatrics [3]. The majority of ADRs with the highest ROR that we found in this study are also described in literature and SPCs (e.g. paradoxical bronchospasm, adrenal insuffi ciency and Churg Strauss syndrome) [13-16]. For example, adrenal insuffi ciency is a well-known side-eff ect of steroids (both inhaled and oral) [15]. Another example is the increased risk of serious adverse drug reactions when using single LABA. A Cochrane review found that serious adverse events were

100

EEliflif FatmaFatma SSenen BW.inddBW.indd 110000 003-01-113-01-11 15:1715:17 Chapter 3.1: Asthma medicines and reported ADRs in children

signifi cantly higher in children using LABA [17]. In 2006, the SMART study showed an excess risk of asthma-related deaths in adult patients using LABA [6]. As a result, the Food and Drug Administration (FDA) issued a ‘black box’ warning for LABA, warning patients about a fatal outcome and recommending never to use LABAs as monotherapy but always to combine it with ICS. Recently, the FDA added further safety requirements to the use of LABAs [18]. For children and adolescent, the FDA requires that when LABA use is required, this should only be a combination product containing both an inhaled corticosteroid and a LABA, to ensure compliance with both medications. ADRs that were not yet described in literature nor mentioned in the SPC were the following: intestinal gangrene, gastric ulcer perforation, heart valve disorders, hemiplegia, tooth disorder, hypokinesia and iritis. Intestinal gangrene, gastric ulcer perforation and heart valve disoders were reported for children using SABAs; hemiplegia and tooth disorders for LABAs, hypokinesia for xanthines and iritis for LTRAs. Most of these ADRs are serious, and in light of the widespread use of asthma medicines in children with asthma, more research is needed to confi rm our fi ndings eventually by means of large observational safety studies. As for all observational research, our data has limitations. First of all, we used data from spontaneous reporting systems, which suff er from bias such as under- and overreporting. [19-21]. Furthermore, it has to be underlined that a signal with a high ROR does not prove that the drug and ADR are causally related. The observed signals are purely hypothesis generating and further research is needed to assess these signals in an appropriate setting [22]. Also, some of the found ADRs are most likely caused by the underlying condition and treatment failure rather than the investigated drugs. This is most probably the case for asthma being reported as an ADR, since all analysed medicines are indicated for use in asthma. Also respiratory acidosis reported for ACH is most probably caused by confounding or treatment failure. Studies have shown that metabolic acidosis occurs in patients with severe acute asthma attacks [18]. Similar explanations hold for rhinitis as ADR for antiallergics, sinces antiallergics are also used to treat rhinitis [23]. Due to the nature of the database, information on indication is missing, and therefore it is not possible to identify the underlying condition. The strength of this study is the large number of pediatric ADRs that was available for analysis. To our knowledge, this is the fi rst study to systematically analyse the spontaneous reports from the Vigibase database for associations between asthma medicines and ADRs in children. It is reassuring to see that the majority of strongly associated ADRs are known already and have described in literature and included in the SPC. On the other hand we found some new signals regarding events that have not yet been described in literature and more research to explore the potential causality is needed.

101

EEliflif FatmaFatma SSenen BW.inddBW.indd 110101 003-01-113-01-11 15:1715:17 References

1 Sturkenboom MC, Verhamme KM, Nicolosi A, Murray ML, Neubert A, Caudri D, et al. Drug use in children: cohort study in three European countries. BMJ (Clinical research ed. 2008;337:a2245. 2 Bianchi M, Clavenna A, Bonati M. Inter-country variations in anti-asthmatic drug prescriptions for children. Systematic review of studies published during the 2000-2009 period. European journal of clinical pharmacology. Sep;66(9):929-36. 3 Sen EF, Verhamme KM, Neubert A, Hsia Y, Murray M, Felisi M, et al. Assessment of Pediatric asthma drug use in three European countries; a TEDDY study. European journal of pediatrics. Sep 2. 4 Graham LM. Balancing safety and effi cacy in the treatment of pediatric asthma. The Journal of allergy and clinical immunology. 2002 Jun;109(6 Suppl):S560-6. 5 Masoli M, Fabian D, Holt S, Beasley R. The global burden of asthma: executive summary of the GINA Dissemination Committee report. Allergy. 2004 May;59(5):469-78. 6 Nelson HS, Weiss ST, Bleecker ER, Yancey SW, Dorinsky PM. The Salmeterol Multicenter Asthma Research Trial: a comparison of usual pharmacotherapy for asthma or usual pharmacotherapy plus salmeterol. Chest. 2006 Jan;129(1):15-26. 7 GINA Report, Global Strategy for Asthma Management and Prevention. [cited; Available from: http:// www.ginasthma.com/GuidelinesResources.asp?l1=2&l2=0 8 European Medicines Agency, Paediatric needs. [cited; Available from: http://www.ema.europa. eu/ema/index.jsp?curl=pages/regulation/document_listing/document_listing_000096. jsp&jsenabled=true 9 Lindquist M. VigiBase, the WHO Global ICSR Database System: Basic Facts. Drug Information Journal. 2008;42:409-19. 10 Methodology. WCCfDS. Guidelines for ATC classifi cation and DDD assignment. [cited; Available from: http://www.whocc.no/ 11 Rothman KJ, Lanes S, Sacks ST. The reporting odds ratio and its advantages over the proportional reporting ratio. Pharmacoepidemiology and drug safety. 2004 Aug;13(8):519-23. 12 INTERNATIONAL CONFERENCE ON HARMONISATION OF TECHNICAL REQUIREMENTS FOR REGISTRA- TION OF PHARMACEUTICALS FOR HUMAN USE. CLINICAL INVESTIGATION OF MEDICINAL PRODUCTS IN THE PEDIATRIC POPULATION E11. ICH Harmonised Tripartite Guideline 2000 [cited 08-Apr-2010]; Available from: 13 Electronic Medicines Compendium. Summary of Product Characteristics for Pulvinal Salbutamol Inhaler 200 micrograms, last updated on the eMC: 26/02/2009. 2010 [cited; Available from: http:// www.medicines.org.uk/EMC/medicine/21311/SPC/Pulvinal+Salbutamol+Inhaler+200+micrograms/ 14 Electronic Medicines Compendium. Summary of Product Characteristics for Bricanyl Respules 2.5 mg/ml Nebuliser Solution, last updated on the eMC: 29/01/2010. 2010 [cited; Available from: http:// www.medicines.org.uk/EMC/medicine/156/SPC/Bricanyl++Respules+2.5+mg+ml+Nebuliser+Solut ion/ 15 Electronic Medicines Compendium. Summary of Product Characteristics for Easyhaler Budesonide 400mcg, last updated on the eMC: 02/07/2009. 2010 [cited; Available from: http://www.medicines. org.uk/EMC/medicine/18774/SPC/Easyhaler+Budesonide+400mcg/ 16 Electronic Medicines Compendium. Summary of Product Characteristics for SINGULAIR PAEDIATRIC 4 mg GRANULES, last updated on the eMC: 02/07/2010. 2010 [cited; Available from: http://www. medicines.org.uk/EMC/medicine/14071/SPC/SINGULAIR+PAEDIATRIC+4+mg+GRANULES/

102

EEliflif FatmaFatma SSenen BW.inddBW.indd 110202 003-01-113-01-11 15:1715:17 Chapter 3.1: Asthma medicines and reported ADRs in children

17 Walters EH, Gibson PG, Lasserson TJ, Walters JA. Long-acting beta2-agonists for chronic asthma in adults and children where background therapy contains varied or no inhaled corticosteroid. Cochrane database of systematic reviews (Online). 2007(1):CD001385. 18 US Food and Drug Administration. FDA Drug Safety Communication: New Safety Requirements for Long-Acting Inhaled Asthma Medications Called Long-Acting Beta-Agonists (LABAs). 18-02-2010 [cited; Available from: http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationfor- PatientsandProviders/ucm200776.htm 19 Hartnell NR, Wilson JP. Replication of the Weber eff ect using postmarketing adverse event reports voluntarily submitted to the United States Food and Drug Administration. Pharmacotherapy. 2004 Jun;24(6):743-9. 20 Motola D, Vargiu A, Leone R, Conforti A, Moretti U, Vaccheri A, et al. Infl uence of regulatory measures on the rate of spontaneous adverse drug reaction reporting in Italy. Drug Saf. 2008;31(7):609-16. 21 Pariente A, Gregoire F, Fourrier-Reglat A, Haramburu F, Moore N. Impact of safety alerts on measures of disproportionality in spontaneous reporting databases: the notoriety bias. Drug Saf. 2007;30(10):891- 8. 22 Almenoff J, Tonning JM, Gould AL, Szarfman A, Hauben M, Ouellet-Hellstrom R, et al. Perspectives on the use of data mining in pharmaco-vigilance. Drug Saf. 2005;28(11):981-1007. 23 Electronic Medicines Compendium. Summary of Product Characteristics for Rynacrom 4% Nasal Spray, last updated on the eMC: 12/09/2007. 2010 [cited; Available from: http://www.medicines.org. uk/EMC/medicine/16124/SPC/Rynacrom+4%25+Nasal+Spray/ 24 Electronic Medicines Compendium. Summary of Product Characteristics for Atrovent Aerocaps, last updated on the eMC: 04/07/2007. 2010 [cited; Available from: http://www.medicines.org.uk/EMC/ medicine/274/SPC/Atrovent+Aerocaps/ 25 Electronic Medicines Compendium. Summary of Product Characteristics for Flixonase Allergy Nasal Spray, last updated on the eMC: 18/03/2009. 2010 [cited; Available from: http://www.medicines.org. uk/EMC/medicine/13481/SPC/Flixonase+Allergy+Nasal+Spray/ 26 Meert KL, Clark J, Sarnaik AP. Metabolic acidosis as an underlying mechanism of respiratory distress in children with severe acute asthma. Pediatr Crit Care Med. 2007 Nov;8(6):519-23. 27 Electronic Medicines Compendium. Summary of Product Characteristics for Intal CFC-free Inhaler, last updated on the eMC: 22/02/2009. 2010 [cited; Available from: http://www.medicines.org.uk/ EMC/medicine/21017/SPC/Intal+CFC-free+Inhaler/ 28 Bernard S. Severe lactic acidosis following theophylline overdose. Annals of emergency medicine. 1991 Oct;20(10):1135-7. 29 Electronic Medicines Compendium. Summary of Product Characteristics for Uniphyllin Continus tablets, last updated on the eMC: 11/03/2010. 2010 [cited; Available from: http://www.medicines.org. uk/EMC/medicine/1233/SPC/Uniphyllin+Continus+tablets/ 30 Electronic Medicines Compendium. Summary of Product Characteristics for Singulair 10 mg Tablets, last updated on the eMC: Singulair 10 mg Tablets. 2010 [cited; Available from: http://www.medicines. org.uk/EMC/medicine/17718/SPC/Singulair+10+mg+Tablets/ 31 Electronic Medicines Compendium. Summary of Product Characteristics for Symbicort Turbohaler 400/12, Inhalation powder, last updated on the eMC: 05/05/2010. 2010 [cited; Available from: http:// www.medicines.org.uk/EMC/medicine/11882/SPC/Symbicort+Turbohaler+400+12%2c+Inhalation +powder./ 32 Electronic Medicines Compendium. Summary of Product Characteristics for Duovent UDVs, last updated on the eMC: 18/06/2009. 2010 [cited; Available from: http://www.medicines.org.uk/EMC/ medicine/6562/SPC/Duovent+UDVs/

103

EEliflif FatmaFatma SSenen BW.inddBW.indd 110303 003-01-113-01-11 15:1715:17 33 Electronic Medicines Compendium. Summary of Product Characteristics for Combivent Metered Aerosol, last updated on the eMC: 17/03/2009. 2010 [cited; Available from: http://www.medicines. org.uk/EMC/medicine/286/SPC/Combivent+Metered+Aerosol/ 34 Gluckman L. Letter: Ventolin psychosis. The New Zealand medical journal. 1974 Nov 13;80(527):411. 35 Vendrell M, de Gracia J, Olveira C, Martinez MA, Giron R, Maiz L, et al. [Diagnosis and treatment of bronchiectasis. Spanish Society of Pneumology and Thoracic Surgery]. Archivos de bronconeumolo- gia. 2008 Nov;44(11):629-40. 36 Electronic Medicines Compendium. Summary of Product Characteristics for Easyhaler Salbutamol 200mcg, last updated on the eMC: 19/03/2010. 2010 [cited; Available from: http://www.medicines. org.uk/EMC/medicine/21085/SPC/Easyhaler+Salbutamol+200mcg/ 37 Lang DM, Davis RS. The long-acting beta-agonist controversy: a clinical dilemma. Allergy Asthma Proc. 2007 Mar-Apr;28(2):136-44. 38 de Vries TW, de Langen-Wouterse JJ, de Jong-Van den Berg LT, Duiverman EJ. Hypertrichosis as a side eff ect of inhaled steroids in children. Pediatric pulmonology. 2007 Apr;42(4):370-3. 39 Choong K, Zwaigenbaum L, Onyett H. Severe varicella after low dose inhaled corticosteroids. The Pediatric infectious disease journal. 1995 Sep;14(9):809-11. 40 New York Allergy & Sinus Centers, Cromolyn Generic Nebulization Solution. [cited; Available from: http://www.nyallergy.com/medications.php?sid=cromolyn 41 Sarkar M, Koren G, Kalra S, Ying A, Smorlesi C, De Santis M, et al. Montelukast use during pregnancy: a multicentre, prospective, comparative study of infant outcomes. European journal of clinical pharmacology. 2009 Aug 26. 42 Electronic Medicines Compendium. Summary of Product Characteristics for Fostair 100/6 inhalation solution, last updated on the eMC: 04/08/2010. 2010 [cited; Available from: http://www.medicines. org.uk/EMC/medicine/21006/SPC/Fostair+100+6+inhalation+solution/

104

EEliflif FatmaFatma SSenen BW.inddBW.indd 110404 003-01-113-01-11 15:1715:17 Chapter 3.2

Cough and cold medicines related adverse drug reactions in children: Results from the WHO adverse events database

E. Fatma Sen, Katia MC Verhamme, Sandra de Bie, Bruno HCh Stricker, Carlo Giaquinto, Adriana Ceci, and Miriam CJM Sturkenboom, on behalf of the TEDDY European Network of Excellence.

EEliflif FatmaFatma SSenen BW.inddBW.indd 110505 003-01-113-01-11 15:1715:17 Abstract

Objective: To provide an overview of pediatric adverse drug reactions (ADR) associated with the use of cough and cold medicines (CCMs) reported to the WHO adverse events reporting database and possibly identify new ADRs associated with CCMs use. Design: case non-case study Setting: worldwide spontaneous reporting database: Vigibase Participants: children aged 0-18 years Main outcome measure: association between cough and cold medicines and suspected ADRs using the reporting odds ratio (ROR) as a measure of disproportionality while adjusting for gender, continent and reporter. To reduce confounding, vaccine-related records were excluded from analysis. Results: Records for CCMs related ADRs comprised 1.2% of the non-vaccine related records (3,467 records). The majority of the reports were for children aged 2-18 years (80%) and most reports concerned mucolytics and opium alkaloids. Severe ADRs such as subarachnoid haemorrhage, myocardial infarction and cerebrovascular disorders were amongst the ADRs with the highest RORs. Most of the found CCMs-ADR associations are described in literature (such as dosing errors, neurologic ADRs associated with opioids use). However, we found some new CCMs-ADR associations such as chronic renal failure associated with use of expectorants (ROR 22.9, 95% CI [7.2-72]) and cerebral infarction associated with the use of opium alkaloids+expectorants (ROR 50.4, 95% CI [23-109]). Conclusion: this overview of pediatric ADRs associated with CCMs use shows that severe and potentially life-threatening ADRs occur in children of all ages. We underline the advice not to use CCMs in young children, and recommend that CCMs, in view of their poor benefi t-risk profi le should not be given to children over all.

106

EEliflif FatmaFatma SSenen BW.inddBW.indd 110606 003-01-113-01-11 15:1715:17 Chapter 3.2: Cough and cold medicines and reported ADRs in children

Introduction

Cough and cold medicines (CCMs) are frequently used to treat upper respiratory tract infections in children. In the past years, safety issues have been raised concerning the use of these drugs in young children. A growing number of studies associated the use of CCMs to serious adverse drug reactions (ADRs) in children, such as cardiac arrhythmias, depressed levels of consciousness and even death [1]. As a result, some countries undertook regulatory actions against over-the-counter use of CCMs in young children [2-4]. Most of the published data on the safety of CCMs deal with children <2 years, however concerns also exist about the safety of these drugs in older children. Because of these concerns, the safety monitoring of of CCMs in older children is ongoing [3]. It is diffi cult to assess the safety of CCMs via randomized controlled trial fi rst of all because very few RCTs have been done in children and the available RCTs have small sample sizes and short follow-up. Spontaneous adverse event reporting databases, such as the Vigibase database of the WHO-UMC (Uppsala Monitoring Centre), can be a source get a better insight in the potential risks of these drugs. To get a broader understanding of the adverse drug reactions associated with CCMs use in children, we analysed the pediatric adverse drug reactions in relation to CCMs reported to the Vigibase database of the WHO-UMC. Our objective was to analyse and describe: 1) pediatric ADRs with the highest reporting odds ratios related to CCMs and 2) possibly identify new safety signals (CCMs-ADR combinations that have not yet been described in literature or the summary of product characteristics)

Methods

107

EEliflif FatmaFatma SSenen BW.inddBW.indd 110707 003-01-113-01-11 15:1715:17 Drugs were additionally coded using the Anatomical Therapeutic Chemical (ATC) classifi cation system of the WHO Collaborating Centre for Drug Statistics Methodology [6]. ATC codes of cough and cold medicines were manually checked for accuracy allocated to the following subgroups: expectorants (e.g. guaifenesin, althea root etc.), mucolytics (e.g. acetylcysteine, ambroxol etc.), opium alkaloids and derivatives (e.g. codeine, dextrometorphan), other cough suppressants (e.g. pentoxyverine, levodropropizine), combination products of opium alkaloids and expectorants, combinations products of other cough suppressants and expectorants, other cold combination preparations, combination products of opium alkaloids and antihistamines (e.g. dextromethorphan+chlorphenamine), combination products of opium alkaloids and sympathomimetics (e.g. dextromethorphan+pseudoephedrine), combination products of other cough suppressants and antihistamines (e.g. guaifenesin+diphenhydramine), combinations products of other cough suppressants and sympathomimetics (e.g. phenobarbital+ephedrine).

Selection of ICSRs From Vigibase all ICSRs in children (age 0 - ≤18 years) received or with onset date between Janu- ary 2000 and December 2006 were extracted. Only spontaneously reported ICSRs and those with a suspected causal relationship between ADR and drug were included. In addition, those ICSRs where information on the reported drug or reported ADR was missing were excluded.

Record defi nition: An ICSR can contain more than one suspected drug, and more than one ADR. We defi ned a record as a unique combination of a drug and an ADR. Hence, an ICSR containing two ADRs with one suspected drug counted for two records, and an ICSR containing two ADRs with two suspected drugs counted for four records. Information on these records include country of origin, reporter, age at onset, year of onset, gender, reported drug, reported ADR, starting and stopping date of the drug, starting and stopping date of the ADR, dosing regimen of the drug, administration route, and causality assessment.

Statistical analysis: All analyses were conducted on a record basis, thus using unique drug-ADR combinations. Disproportionality was assessed via the calculation of the reporting odds ratio (ROR), which is calculated analogues to and odds ratio [7]. RORs with 95% confi dence intervals (95% CI) were calculated for the diff erent subgroups of cough and cold medicines. The ROR was only calculated if at least 3 records of the drug-ADR combination of interest were available. A ROR of ≥ 2 with a 95% confi dence interval not including 1 is considered signifi cant. Since this method is purely hypothesis generating it must be emphasized that a signifi cant signal does not imply causality. To study the eff ects of age; the ROR was calculated within predefi ned age-categories. In line with the guidelines of the International Conference of Harmonization (ICH), age at the time

108

EEliflif FatmaFatma SSenen BW.inddBW.indd 110808 003-01-113-01-11 15:1715:17 Chapter 3.2: Cough and cold medicines and reported ADRs in children

of the event was categorized into the following age categories: 0 - < 2 years, 2 - ≤11 years and 12 - ≤18 years [8]. Since vaccines represent a substantial part of the drugs used in children and may bias the RORs of other drug-ADR combinations, vaccine related records were excluded.. Additional logistic regression analysis using SPSS version 15.0 for Windows (SPSS, Inc., Chicago, Illinois) was performed to correct for gender, continent and type of reporter.

Results

In the period between January 2000 and December 2006 221,508 suspected (thus potentially causally related) spontaneous ICSRs in children aged ≤18 years were reported to the WHO- UMC, containing a total of 845,090 records. After exclusion of records that were vaccine related (n= 517,642) and for which information on the reported drug or reported ADR was missing (n= 32,675), the total number of drug-ADR combinations in the analysis was 294,773. General characteristics of all records and ATC R05 related records are presented in table 1. The database comprised 3,467 CCMs related records (1.2% of total records). Compared to all drugs (including records of CCMs), records of CCMs related ADRs involved more females, were more often reported by a physician, were less likely to originate from the United States and involved younger children (mean age diff erence of 1 year).

Table 1: Characteristics of all records and cough and cold medicines (CCM) related records CCM related All records % records % N 294,773 3,467 Gender; male 153,708 52.1 1,628 47.0 Notifi er General practitioner 53,790 18.2 456 13.2 Physician 50,300 17.1 759 21.9 Hospital 38,795 13.2 456 13.2 Other 151,888 51.5 1,796 51.8 Country of origin United States 181,126 61.4 1,470 42.4 France 18,820 6.4 555 16.0 Thailand 16,464 5.6 170 4.9 Age Mean / SD 9.7 6.0 8.7 6.3 Median 10.0 8.0 0 - < 2 years 45,209 15.3 601 17.3 2 - ≤ 11 years 114,704 38.9 1,477 42.6 12 - ≤ 18 years 134,860 45.8 1,389 40.1

109

EEliflif FatmaFatma SSenen BW.inddBW.indd 110909 003-01-113-01-11 15:1715:17 Distribution of records by drug class The distribution of CCMs related records by drug class and age category is given in Table 2. Most ADR records involved children aged 2-≤11 years (43%), followed by 12-≤18 years (40%) and children aged <2 years (17%). ADRs were most frequently reported for mucolytics (30.3%) and opium alkaloids and derivatives (20.5%).

Table 2: Distribution of cough and cold preparations related records by drug class and age category Drug class 0-2 years 2-11 years 11-18 Total years N %* N %* N %* N%* Expectorants 40 6.7 103 7.0 48 3.5 191# 7.9 Mucolytics 179 29.8 474 32.1 397 28.6 1,050# 30.3 Opium alkaloids and derivatives 92 15.3 259 17.5 359 25.8 710# 20.5 Other cough suppressants 47 7.8 126 8.5 103 7.4 276# 8 Combinations of opium alkaloids and expectorants 15 2.5 104 7.0 156 11.2 275# 7.9 Combinations of other cough suppressants and expectorants 0 0.0 12 0.8 0 0.0 12# 0.3 Combinations of opium alkaloids and antihistamines 112 18.6 166 11.2 110 7.9 388# 11.2 Combinations of opium alkaloids and sympathicomimetics 80 13.3 149 10.1 160 11.5 389# 11.2 Combinations of other cough suppressants and antihistamines 0 0.0 1 0.1 0 0.0 1 0 Combinations of other cough suppressants and sympathicomimetics 2 0.3 2 0.1 0 0.0 4 0.1 Other cold combination preparations 34 5.7 81 5.5 56 4.0 171# 2.6 Total 601 100 1,477 100 1,389 100 3,467 100 * = % of total within this age category # = statistically signifi cant diff erence (p<0.05) between number of records in the age categories for that drug class

Drug-ADR combinations with the highest RORs The highest ROR between CCMs as a group and specifi c events were found for cystic fi brosis (ROR 64.0, 95% CI 20-203) and hepatic haemorrhage (ROR (54.2, 95% CI 18-167)) both mainly in children aged 12-≤18 years (table 3). Severe ADRs such as subarachnoid haemorrhage and cerebral infarction were also among the ADRs with the highest RORs. Subarachoid haemorrhage was only reported in children aged 12-≤18 years, and cerebral infarction in both children aged 2-≤11 and 12-≤18 years (table 3).

Expectorants: Analyses by type of CCM showed that for expectorants the most disproportionally reported ADRs were subarachnoid haemorrhage, chronic renal failure and cerebrovascular disorder (table 4 and 5). When stratifying by age category, ROR for expectorants was the highest for apnoea in children aged 0-2 years, chronic renal failure in children 2-≤11 years and cerebrovascular ADRs (subarachnoid haemorrhage and cerebrovascular disorder) in the oldest age category.

110

EEliflif FatmaFatma SSenen BW.inddBW.indd 111010 003-01-113-01-11 15:1715:17 Chapter 3.2: Cough and cold medicines and reported ADRs in children

Table 3: Ten highest reporting odds ratios (RORs) for ADRs associated with cough and cold preparations (CCM) stratifi ed by age All age categories Stratifi ed by age category Associated Associated ROR* ROR* ROR* Adverse drug with any ROR crude ROR* with CCM (95% CI) (95% CI) (95% CI) reaction other drug (95% CI) (95% CI) N 0-2 yr 2-≤11 yr 12-≤18 yr N 64.0 2981x103 Cystic Fibrosis 5 7 60.1 (19-189) -- (20-203) (0-infi nity) Hepatic 54.2 5 8 52.6 (17-161) - - 65.2 (21-206) Haemorrhage (18-167) Infusion site 21.0 3 12 21.0 (5.9-75) - - 35.6 (9.4-135) reaction (5.9-75) 13.3 23.3 Haemoptysis 11 70 13.2 (7.0-25) - 12.3 (5.6-27) (7.0-25) (7.5-72) Subarachnoid 13.7 6 39 12.9 (5.5-31) - - 29.5 (9.7-90) Haemorrhage (5.8-32) Incorrect drug 12.4 12.2 administration 3 21 12.0 (3.5-40) -- (3.7-42) (3.6-42) dosage 4.1 Cerebral infarction 13 130 8.4 (4.8-15) 8.6 (4.9-15) - 16.2 (8.3-32) (1.3-13) Cerebrovascular 14.4 8.0 28 284 8.3 (5.7-12) 8.7 (5.9-13) 8.0 (4.4-14) Disorder (6.0-35) (4.2-15) Pleurisy 3 33 7.6 (2.3-25) 6.3 (1.9-21) - - - Hypernatraemia 5 62 6.8 (2.7-17) 6.9 (2.8-17) - - 21.4 (7.2-64) * = adjusted for gender, continent and type of reporter - = number drug-ADR combination with ≥ 3 records

Mucolytics: Mucolytics were strongly associated with an aggravation of the condition, involuntary muscle contractions and angioedema in the children 0-2 years. In older children, haemoptysis, urticaria, meningitis and infusion site reaction were the ADRs with the highest RORs.

Opium alkaloids and derivates Opium alkaloids and derivatives were associated with neurological ADRs in children younger than 12 years namely stupor in children < 2 years and ataxia, neurologic disorder and mydriasis in children aged 2-11 years. In children 12-18 years, pulmonary problems were most strongly associated with opium alkaloids.

Other cough suppressants (e.g. pentoxyverine) The therapeutic class “other cough suppressants” was most strongly associated with coma in children < 2 years. In the category 2-≤11 years, psychiatric ADRs (delirium and hallucination) and in children 12-≤18 years increased sweating, coma and increased levels of SGOT were most strongly associated.

111

EEliflif FatmaFatma SSenen BW.inddBW.indd 111111 003-01-113-01-11 15:1715:17 Table 4: Three highest reporting odds ratios (RORs) by drug class Number Number of ADRs of ADRs Described in Drug class ADR associated associated ROR* (95%CI) literature with drug of with other interest drugs Subarachnoid 4 41 138.6 (49-393) Yes [24] Haemorrhage Expectorants Chronic Renal Failure 3 217 22.9 (7.2-72) No Cerebrovascular 5 307 22.8 (9.3-56) Yes [24] Disorder Confounding Cystic Fibrosis 5 7 216.1 (66-706) by indication Mucolytics Infusion site reaction 3 12 74.8 (-20-275) Yes [25] Confounding Haemoptysis 10 71 39.3 (20-77) by indication Pulmonary 6 133 16.1 (6.6-39) Yes [11] Opium alkaloids and Congestion derivatives Drug Dependence 5 250 9.5 (3.9-23) Yes [11] Ataxia 20 525 9.0 (5.1-16) Yes [11] Delirium 3 206 13.9 (4.4-44) Yes [26] Other cough Oedema cerebral 3 231 13.6 (4.3-43) Yes [27] suppressants Arrhythmia 3 411 7.3 (2.3-23) Yes [28] Cerebral infarction 7 136 50.4 (23-109) No Combinations of Myocardial Infarction 5 115 43.2 (17-107) Yes [24, 29] opium alkaloids and Cerebrovascular expectorants 11 301 32.6 (18-60) Yes [24, 30] Disorder Incorrect drug administration 3 21 121.6 (36-411) Yes [31, 32] Combinations of dosage opium alkaloids and Bleeding time antihistamines 3 42 56.9 (18-185) Yes [33] increased Hepatic Necrosis 3 106 22.9 (7.2-73) Yes [34, 35] Combinations of Faeces Discoloured 3 87 25.0 (7.9-80) No opium alkaloids and Heart Disorder 4 132 19.8 (7.3-54) Yes [3, 36] sympathomimetics Allergy 3 119 21.4 (6.8-68) Yes [37, 38] Angioedema 3 1324 6.3 (2.0-20) Yes [39, 40] Other cold Urticaria 10 7443 5.2 (2.6-10) Yes [39, 40] combination products Face Oedema 3 2156 5.2 (1.7-17) Yes [40] * = adjusted for gender, continent and type of reporter

Combination products Combination products of opium alkaloids and expectorants were strongly associated with neurological disorders in each age category within the children. In the category 2-≤11 years allergic reaction, and in children 12-≤18 heart disorders and myocardial infarction were strongly associated as well. Combinations of opium alkaloids and antihistamines were mostly associated with liver injury related ADRs in children <2 years, such as an increased bleeding time, hyperammonaemia and

112

EEliflif FatmaFatma SSenen BW.inddBW.indd 111212 003-01-113-01-11 15:1715:17 Chapter 3.2: Cough and cold medicines and reported ADRs in children

hepatic necrosis. In children aged 2-≤11 years, neurologic disorder were amongst the most strongly associated ADRs (hypertonia, mydriasis). In the older age category, mainly psychiatric andneurologic ADRs were found to have the highest RORs (paranoid reaction, apathy and abnormal thinking). For the combinations of opium alkaloids and sympathomimetics the highest RORs were found for neurologic ADRs in children <2 years. In children 2-≤11 years, hallucination was most strongly associated to these drugs, whereas in children 12-≤18 years heart disorder was the ADR with the highest ROR. ADRs for other cold combination products were only reported for the age categories 0-2 and 2-≤11 years. Allergic reactions such as urticaria and angioedema were ADRs with the highest ROR in both children aged 0-2 and 2-≤11 years. For the drug groups combinations of other cough suppressants and expectorants, combinations of other cough suppressants and antihistamines, and combinations of other cough suppressants and sympathomimetics, no drug-ADR combination with at least 3 records were reported to the database.

Discussion

This is the fi rst study that analysed pediatric ADRs reported for cough and cold medicines (CCMs) in the Vigibase WHO-UMC database. There are several important fi ndings in our study. First, CCMs related ADRs were infrequently reported (1.2% of the non-vaccine related records) although studies have shown that in a given week, 10% of the children use a CCM [9]. The low reporting rate could be due to the fact that most CCMs are sold over the counter and therefore monitoring is lacking. Second, most CCMs related ADRs were reported for children aged 2-≤11 and 12-18 years of age although utilization studies have shown that prevalence of use of CCMs is highest in younger children and decreases with age [9]. There thus seems to be a discrepancy between the percentage of CCMs use in children <2 and the proportion of ADRs reported for this age group. A possible explanation could be that ADRs are not recognized in very young children. Third, most CCMs related ADRs concerned mucolytics and opium alkaloids, although these are not the most commonly used CCMs in pediatrics [9]. Forth, among the ADRs with the highest RORs, serious and potentially life threatening reactions such as hepatic haemorrhage, subarachnoid haemorrhage and cerebral infarction were reported. Our results are in line with the existing literature studying CCMs-related ADRs. Most of the ADRs that we found in this study, are also described in literature (Tables 4 & 5). For example, in our analyses, dosing errors are amongst the ADRs with the highest RORs in children aged 0-2 years. A recent study conducted by Dart et al. concluded that 85 % of the pediatric fatalities

113

EEliflif FatmaFatma SSenen BW.inddBW.indd 111313 003-01-113-01-11 15:1715:17 Yes [25] Yes Yes [24] Yes Yes [24] Yes by indication by by indication by Confounding Confounding Confounding Confounding Described in literature

4 nity) 516) 163) 1654) 50.0 (15- 1026x10 131.2 (33- (0-infi 8.6 (3.5-21) [11] Yes 8.8 (2.8-28) No 453.9 (125- 25.2 (10-63) [11] Yes ROR* (95%CI) ADR Pulmonary Oedema Sweating Increased Haemoptysis 35.0 (15-82) Pulmonary Congestion Infusion site reaction Cystic Fibrosis Cystic Cerebrovascular Cerebrovascular Disorder Subarachnoid Haemorrhage No Yes [50]Yes Coma [51]Yes Increased SGOT 5.1 (1.6-16) 8.6 (3.8-20) No [47] Yes Yes [48] Yes Yes [11] Yes Yes [11]Yes Medication Error 10.5 (7.0-16) [31] Yes Yes [11] Yes Yes [46] Yes Yes [43]Yes Asthenia 14.3 (4.4-46) [43] Yes Yes [43] Yes indication meningitis) No (mogelijk confounding, confounding, rinosinusitis-> Confounding by by Confounding Described in literature 11) 15) 28) 24) 21) 26) 31) 26) 9.7) 105) 268) 114) 3.6 (1.3- 3.5 (1.1- 5.7 (2.1- 8.7 (2.8- 8.8 (3.2- 9.7 (4.5- 9.7 (3.0- 9.6 (3.5- 82.6 (25- 35.3 (11- 32.1 (9.8- 10.5 (4.2- ROR* (95%CI) ADR Hallucination Neurologic Neurologic NOS Disorder Chronic Renail Chronic Failure [41] Yes [49]Yes Angioedema Yes [47]Yes Delirium Yes [11]Yes Mydriasis Yes [11] Yes Yes [46]Yes Meningitis [11]Yes Ataxia Yes [45]Yes Urticaria Acute Yes [42]Yes Pneumonia [44]Yes Haemoptysis Yes [42]Yes Pain Yes, in animals Yes, Described in literature age categoryed by 23) 49) 17) 14) 19) 64) 15) 11) 57) 15) 33) 8.9 (3.5- 6.2 (2.3- 6.8 (3.3- 5.9 (1.9- 4.6 (1.4- 5.2 (1.8- 36.8 (24- 15.1 (4.6- 19.9 (6.1- 3.4 (1.01- 10.0 (3.1- ROR* (95%CI) 0-2 years 2-11 years 12-18 years ADR Convulsion Coma Cyanosis Stupor Somnolence Angioedema Erythematous Rash aggrevated Condition Apnoea Involuntary Muscle Contractions Maculopapular Rash Drug Class Other cough suppressants Opium alkaloids and Opium alkaloids derivatives Mucolytics Expectorants Table 5: Three highest RORs by drug class stratifi highest RORs by Three 5: Table

114

EEliflif FatmaFatma SSenen BW.inddBW.indd 111414 003-01-113-01-11 15:1715:17 Chapter 3.2: Cough and cold medicines and reported ADRs in children No Yes [24] Yes Yes [24] Yes 114) 111) 169) - 35.2 (11- 39.8 (14- 71.6 (30- 22.2 (8.1-61) [53] Yes Thinking Abnormal Myocardial Myocardial Infarction No Reaction Paranoid 31.1 (11-86) [53] Yes No Heart Disorder No? Mydriasis 9.6 (3.0-30) [11] Yes Yes [10]Yes Arrest Cardiac 5.4 (1.7-17) [36] Yes Yes [55]Yes Heart Disorder 35.2 (13-98) [36] Yes Yes [11] Yes Yes [54]Yes Apathy 28.7 (12-71) [11, 53] Yes Yes [52] Yes Yes [57] Yes Yes [24]Yes Infarction Cerebral Yes [39, 40] Yes Yes [39, 40] Yes 29) 18) 15) 28) 43) 43) 13) 8.2) 8.7) -19) 141) 103) 7.8 (3.1 4.0 (1.2- 4.5 (2.5- 8.9 (2.7- 6.6 (2.4- 7.9 (4.0- 3.4 (1.3- 407 (16- 43.2 (13- 11.3 (4.6- 15.8 (5.8- 17.6 (7.2- Medication Error Therapeutic Therapeutic Response increased Acidosis Allergic Reaction Allergic Urticaria Cerebrovascular Cerebrovascular Disorder Hydrocephalus Yes [56]Yes Angioedema Yes [11] Yes Yes [11]Yes Hallucination Yes [33]Yes Hypertonia Yes [11] Yes Yes [39, 40]Yes Somnolence Yes [34, 35]Yes Mydriasis by indication by Confounding Confounding er 33) 40) 28) 57) 9.1) 199) 357) 186.2 124.5 9.7 (2.8- 3.2 (1.2- (44-782) (33-463) 55.9 (16- 95.1 (25- 13.7 (4.7- 10.1 (3.7- 17.5 (5.3- 0-2 years 2-11 years 12-18 years Urticaria Convulsions Coma Neurologic Disorder Neurologic Coughing Hepatic Necrosis Hyperammonaemia Bleeding time increased Neurologic Disorder Neurologic Drug Class Other cold products combination Combinations of Combinations and opium alkaloids sympathomimetics Combinations of Combinations and opium alkaloids antihistamines Combinations of Combinations opium alkaloids and expectorants Table 5: Continue Table and type continent of notifi gender, for * = adjusted with ≥ 3 records - = number of drug-ADR combination

115

EEliflif FatmaFatma SSenen BW.inddBW.indd 111515 003-01-113-01-11 15:1715:17 associated with CCMs use, was due to an overdose of CCMs, mainly in children under the age of 2 [10]. Another example of known ADRs found in this study are the typical (of characteristic) adverse eff ects associated with the use of opium alkaloids, such as stupor, ataxia and mydriasis [11]. Very few ADRs were not yet described in literature namely chronic renal failure associated with expectorants use, cerebral infarction associated with opium alkaloids+expectorants use and discoloration of the faeces associated with use of opium alkaloids+antihistamines. As Chronic renal failure and cerebral infarction are serious and potentially life-threatening adverse events, more research is needed to confi rm our fi ndings eventually by means of large observational safety studies. . When stratifying by age category, no clear pattern in type of ADR and severity could be observed. For all all age groups, various ADRs of diff erent levels of severity were reported. Over the counter use of CCMs in children <6 years has been banned in many countries, because of safety concerns [2-4]. Safety issues are probably not only limited to younger children and use of CCMs in this age category is substantial. This explains why the safety of these drugs in older children is currently under review by the FDA [3, 9, 12]. Lookin at our data, we also found several severe ADRs (e.g. subarachnoid haemorrhage, hydrocephalus, hepatic haemorrhage) associated with CCMs both in children aged 2-≤11 and 12-≤18 years. The fact that there is no evidence of effi cacy of CCMs in children and our observed fi ndings underline the need for further safety data of CCMs both in young and older children [13-16]. As for all observational research, our data has limitations. First of all, we used data from spontaneous reporting systems, which suff er from bias such as under and overreporting. [20-22]. Furthermore, is has to be underlined that a signal with a high ROR does not prove that the drug and ADR are causally related. The observed signals are purely hypothesis generating and further research is needed to assess these signals in an appropriate setting [23]. Also, some of the found ADRs are most likely caused by the underlying condition, rather than the investigated drugs (confounding by indication). This is probably the case for the ADRs cystic fi brosis, haemoptysis and meningitis reported with the use of mucolytics. The mucolytic agent dornase alfa is indicated for use in patients with cystic fi brosis [17]. Mucolytics are also used to treat cough, which can cause haemoptysis. Meningitis is most probably a complication of bacterial rhinosinusitis, for which mucolytics could be given. It is described in literature that meningitis can be a severe complication in patients with bacterial rhinosinusitis [18]. Hydrocephalus associated with use of opium alkaloids+expectorants could also be confounding, since the cough for which these drugs are given for, can be the fi rst symptom of hydrocephalus [19]. However, hydrocephalus has also been described as a complication of methadon (an opioid) use. Since opium alkaloids and methadon have the same pharmacological origin, hydrocephalus could be caused by the use of opium alkaloids+expectorants.

116

EEliflif FatmaFatma SSenen BW.inddBW.indd 111616 003-01-113-01-11 15:1715:17 Chapter 3.2: Cough and cold medicines and reported ADRs in children

Due to the nature of the database, information on indication is missing, and therefore it is not possible to identify the underlying condition. The strength of this study is the large number of pediatric ADRs available for analysis. To our knowledge, this is the fi rst study analysing the spontaneous reports from the Vigibase database for associations between CCMs and pediatrics ADRs. We found some new ADRs associated with some drug groups which have not yet been described in literature and more research to explore the potential causality is needed. . Our results show that serious ADRs associated with CCMs occur in children of all ages. We underline the advice not to use CCMs in young children, and recommend that CCMs should also not be used in pediatrics at all.

117

EEliflif FatmaFatma SSenen BW.inddBW.indd 111717 003-01-113-01-11 15:1715:17 None of the authors has a confl ict of interest. The principal investigator had full access to all of the data in the study and takes responsibility for their integrity and the accuracy of the data analysis.

Funding Funded by the European Community’s 6th Framework Programme. Project number LSHB- CT-2005-005216: TEDDY: Task force in Europe for Drug Development for the Young.

Disclaimer: Data from the WHO Collaborating Centre for International Drug Monitoring was used. The information is not homogeneous at least with respect to origin or likelihood that the pharmaceutical product caused the adverse reaction. The information does not represent the opinion of the WHO.

118

EEliflif FatmaFatma SSenen BW.inddBW.indd 111818 003-01-113-01-11 15:1715:17 Chapter 3.2: Cough and cold medicines and reported ADRs in children

References

1 Sharfstein JM, North M, Serwint JR. Over the counter but no longer under the radar--pediatric cough and cold medications. The New England journal of medicine. 2007 Dec 6;357(23):2321-4. 2 Canada H. Health Canada Advisory: Cough and Cold products for children. 2009 [cited 07-10-2009]; Available from: http://chd.region.waterloo.on.ca/web/health.nsf/vwSiteMap/E7742A6975D6895E85 2570E50055EEFF/$fi le/PHYSUP_JAN09.pdf?openelement 3 FDA. FDA Recommends that Over-the-Counter (OTC) Cough and Cold Products not be used for Infants and Children under 2 Years of Age. 2008 [cited 03-10-2009]; Available from: http://www.fda. gov/Drugs/DrugSafety/PublicHealthAdvisories/ucm051137.html 4 MHRA. Press release: Better medicines for children’s coughs and colds. 2009 [cited 03-10-2009]; Avail- able from: http://www.mhra.gov.uk/NewsCentre/Pressreleases/CON038902 5 Lindquist M. VigiBase, the WHO Global ICSR Database System: Basic Facts. Drug Information Journal. 2008;42:409-19. 6 Methodology. WCCfDS. Guidelines for ATC classifi cation and DDD assignment. [cited; Available from: http://www.whocc.no/ 7 Rothman KJ, Lanes S, Sacks ST. The reporting odds ratio and its advantages over the proportional reporting ratio. Pharmacoepidemiology and drug safety. 2004 Aug;13(8):519-23. 8 INTERNATIONAL CONFERENCE ON HARMONISATION OF TECHNICAL REQUIREMENTS FOR REGISTRA- TION OF PHARMACEUTICALS FOR HUMAN USE. CLINICAL INVESTIGATION OF MEDICINAL PRODUCTS IN THE PEDIATRIC POPULATION E11. ICH Harmonised Tripartite Guideline 2000 [cited 08-Apr-2010]; Available from: 9 Vernacchio L, Kelly JP, Kaufman DW, Mitchell AA. Cough and cold medication use by US children, 1999-2006: results from the slone survey. Pediatrics. 2008 Aug;122(2):e323-9. 10 Dart RC, Paul IM, Bond GR, Winston DC, Manoguerra AS, Palmer RB, et al. Pediatric fatalities associated with over the counter (nonprescription) cough and cold medications. Annals of emergency medicine. 2009 Apr;53(4):411-7. 11 Olson KR. Poisoning & Drug Overdose, Fourth Edition. 2004 [cited May 26, 2010]; Available from: http://www.hrd.gov.co/INTOXICACION/Poisoning%20and%20Drug%20Overdose.pdf 12 Gadomski AM, Rubin JD. Cough and cold medicine use in young children: a survey of Maryland pediatricians. Md Med J. 1993 Jul;42(7):647-50. 13 Infant deaths associated with cough and cold medications--two states, 2005. Mmwr. 2007 Jan 12;56(1):1-4. 14 Rimsza ME, Newberry S. Unexpected infant deaths associated with use of cough and cold medications. Pediatrics. 2008 Aug;122(2):e318-22. 15 Wingert WE, Mundy LA, Collins GL, Chmara ES. Possible role of pseudoephedrine and other over-the- counter cold medications in the deaths of very young children. Journal of forensic sciences. 2007 Mar;52(2):487-90. 16 Marinetti L, Lehman L, Casto B, Harshbarger K, Kubiczek P, Davis J. Over-the-counter cold medications-postmortem fi ndings in infants and the relationship to cause of death. Journal of analytical toxicology. 2005 Oct;29(7):738-43. 17 Electronic Medicines Compendium. Summary of Product Characteristics for Pulmozyme 2500 U/ 2.5ml, nebuliser solution, last updated on the eMC: 20/09/2010. 2010 [cited; Available from: http:// www.medicines.org.uk/EMC/medicine/1723/SPC/Pulmozyme+2500+U++2.5ml%2c+nebuliser+sol ution/

119

EEliflif FatmaFatma SSenen BW.inddBW.indd 111919 003-01-113-01-11 15:1715:17 18 Amat F. [Complications of bacterial rhino-sinusitis in children: a case report and a review of the literature]. Arch Pediatr. Mar;17(3):258-62. 19 Karatayli-Ozgursoy S, Dominik J, Eidelman B, Guarderas JC. Chronic cough as the presenting symptom of hydrocephalus. Southern medical journal. Jun;103(6):574-7. 20 Hartnell NR, Wilson JP. Replication of the Weber eff ect using postmarketing adverse event reports voluntarily submitted to the United States Food and Drug Administration. Pharmacotherapy. 2004 Jun;24(6):743-9. 21 Motola D, Vargiu A, Leone R, Conforti A, Moretti U, Vaccheri A, et al. Infl uence of regulatory measures on the rate of spontaneous adverse drug reaction reporting in Italy. Drug Saf. 2008;31(7):609-16. 22 Pariente A, Gregoire F, Fourrier-Reglat A, Haramburu F, Moore N. Impact of safety alerts on measures of disproportionality in spontaneous reporting databases: the notoriety bias. Drug Saf. 2007;30(10):891-8. 23 Almenoff J, Tonning JM, Gould AL, Szarfman A, Hauben M, Ouellet-Hellstrom R, et al. Perspectives on the use of data mining in pharmaco-vigilance. Drug Saf. 2005;28(11):981-1007. 24 Lee L. ODS Postmarketing Safety Review (PID No. D030159) Ipecac Executive Summary Center for Drug Evaluation and Research, FDA. May 6, 2003 [cited; Available from: http://www.fda.gov/ohrms/ dockets/ac/03/briefi ng/3962B1_12_Post-Marketing%20Safety%20Review.pdf 25 Electronic Medicines Compendium. Summary of Product Characteristics for Parvolex Solution for Injection, last updated on the eMC: 05/05/2010. 2010 [cited; Available from: http://www.medicines. org.uk/EMC/medicine/1127/SPC/Parvolex+Solution+for+Injection/ 26 Lauby V, Imbert B, Eysseric H, Dufrene I, Mallaret M. [Neuropsychic side-eff ects of therapeutic doses of zipeprol]]. La Revue de medecine interne / fondee 1995;16(3):212-3. 27 Perraro F, Beorchia A. Convulsions and cerebral oedema associated with zipeprol abuse. Lancet. 1984 Jan 7;1(8367):45-6. 28 Clobutinol: severe cardiac arrhythmias. Old drugs should also be monitored. Prescrire international. 2008 Jun;17(95):113. 29 Backmund M, Meyer K, Zwehl W, Nagengast O, Eichenlaub D. Myocardial Infarction associated with methadone and/or dihydrocodeine. European addiction research. 2001 Mar;7(1):37-9. 30 Klein-Schwartz W, Gorman RL, Oderda GM, Wedin GP, Saggar D. Ipecac use in the elderly: the unan- swered question. Annals of emergency medicine. 1984 Dec;13(12):1152-4. 31 Shah K, Barker KA. Out-of-hospital medication errors: a 6-year analysis of the national poison data system. Pharmacoepidemiology and drug safety. 2009 Nov;18(11):1080-5. 32 Crouch BI, Caravati EM, Moltz E. Tenfold therapeutic dosing errors in young children reported to U.S. poison control centers. Am J Health Syst Pharm. 2009 Jul 15;66(14):1292-6. 33 Electronic Medicines Compendium. Summary of Product Characteristics for Ranitidine 300mg Tab- lets, last updated on the eMC: 07/04/2010. 2010 [cited; Available from: http://www.medicines.org.uk/ EMC/medicine/22915/SPC/Ranitidine+300mg+Tablets/#PRODUCTINFO 34 Fleckenstein JL. Nyquil and acute hepatic necrosis. The New England journal of medicine. 1985 Jul 4;313(1):48. 35 Mercer AE, Regan SL, Hirst CM, Graham EE, Antoine DJ, Benson CA, et al. Functional and toxicological consequences of metabolic bioactivation of methapyrilene via thiophene S-oxidation: Induction of cell defence, apoptosis and hepatic necrosis. Toxicology and applied pharmacology. 2009 Sep 15;239(3):297-305. 36 Gunn VL, Taha SH, Liebelt EL, Serwint JR. Toxicity of over-the-counter cough and cold medications. Pediatrics. 2001 Sep;108(3):E52.

120

EEliflif FatmaFatma SSenen BW.inddBW.indd 112020 003-01-113-01-11 15:1715:17 Chapter 3.2: Cough and cold medicines and reported ADRs in children

37 Electronic Medicines Compendium. Summary of Product Characteristics for ALLERcalm Allergy Relief Tablets, last updated on the eMC: 07/09/2007. 2010 [cited; Available from: http://www.medicines.org. uk/EMC/medicine/20065/SPC/ALLERcalm+Allergy+Relief+Tablets/ 38 Electronic Medicines Compendium. Summary of Product Characteristics for Vicks Cough Syrup with Honey for Dry Coughs, last updated on the eMC: 05/08/2010. 2010 [cited; Available from: http://www. medicines.org.uk/EMC/medicine/20324/SPC/Vicks+Cough+Syrup+with+Honey+for+Dry+Coughs/ 39 Iorio ML, Moretti U, Colcera S, Magro L, Meneghelli I, Motola D, et al. Use and safety profi le of antiepileptic drugs in Italy. European journal of clinical pharmacology. 2007 Apr;63(4):409-15. 40 Summary of Product Characteristics for Phenobarbital 30mg Tablets BP, last revised on 09/01/2007. [cited; Available from: http://www.mhra.gov.uk/home/groups/l-unit1/documents/websiteresources/ con2030939.pdf 41 Kalhoro AB, Rex MA. Observations on the use of glyceryl guaiacolate as an adjunct to general anaes- thesia in horses. Australian veterinary journal. 1984 Feb;61(2):49-53. 42 Electronic Medicines Compendium. Summary of Product Characteristics for Benylin Childrens Chesty Coughs, last updated on the eMC: 17/03/2010. 2010 [cited; Available from: http://www.medicines. org.uk/EMC/medicine/7082/SPC/Benylin+Childrens+Chesty+Coughs/ 43 Silber TJ. Ipecac syrup abuse, morbidity, and mortality: isn’t it time to repeal its over-the-counter status? J Adolesc Health. 2005 Sep;37(3):256-60. 44 Chalumeau M, Cheron G, Assathiany R, Moulin F, Bavoux F, Breart G, et al. [Mucolytic agents for acute respiratory tract infections in infants: a pharmacoepidemiologic problem?]. Arch Pediatr. 2002 Nov;9(11):1128-36. 45 Bailey B, Blais R, Letarte A. Status epilepticus after a massive intravenous N-acetylcysteine overdose leading to intracranial hypertension and death. Annals of emergency medicine. 2004 Oct;44(4):401-6. 46 Flanagan RJ, Meredith TJ. Use of N-acetylcysteine in clinical toxicology. The American journal of medicine. 1991 Sep 30;91(3C):131S-9S. 47 Winter ML, Spiller HA, Griffi th JR. Benzonatate Ingestion Reported to the National Poison Center Database System (NPDS). J Med Toxicol. May 20. 48 Meyboom RH. Adverse reaction to drugs in children, experiences with “spontaneous monitoring” in The Netherlands. Bratislavske lekarske listy. 1991 Nov;92(11):554-9. 49 Merigot P, Garnier R, Efthymiou ML. [Convulsions with 3 antitussive substituted derivatives of pipera- zine (zipeprol, eprazinone, eprozinol)]. Annales de pediatrie. 1985 Jun;32(6):504-6, 11. 50 Seitz CS, Brocker EB, Trautmann A. Allergy evaluation after emergency treatment: anaphylaxis to the over-the-counter medication clobutinol. Emerg Med J. 2007 Mar;24(3):e19. 51 McEwen J, Meyboom RH, Thijs I. Hallucinations in children caused by oxolamine citrate. The Medical journal of Australia. 1989 Apr 17;150(8):449-50, 52. 52 van Ketel WG. Rash from oral codeine. Contact dermatitis. 1986 Sep;15(3):195. 53 Simons FE. Advances in H1-antihistamines. The New England journal of medicine. 2004 Nov 18;351(21):2203-17. 54 Ponsonby AL, Dwyer T, Couper D. Factors related to infant apnoea and cyanosis: a population-based study. Journal of paediatrics and child health. 1997 Aug;33(4):317-23. 55 Sauder KL, Brady WJ, Jr., Hennes H. Visual hallucinations in a toddler: accidental ingestion of a sympathomimetic over-the-counter nasal decongestant. The American journal of emergency medicine. 1997 Sep;15(5):521-6. 56 Ruha AM, Graeme KA, Field A. Late seizure following ingestion of Vicks VapoRub. Acad Emerg Med. 2003 Jun;10(6):691.

121

EEliflif FatmaFatma SSenen BW.inddBW.indd 112121 003-01-113-01-11 15:1715:17 57 Electronic Medicines Compendium. Summary of Product Characteristics for Children’s 6 Months Plus Cough and Congestion Syrup, last updated on the eMC: 27/08/2003. 2010 [cited; Available from: http://www.medicines.org.uk/EMC/medicine/8449/SPC/Children%27s+6+Months+Plus+Cough+an d+Congestion+Syrup/ 58 Food and Drug Administration DoDRE. Nonprescription Drug Advisory Committee meeting: cold, cough, allergy, bronchodilator, antiasthmatic drug products for over-the-counter human use. 2007:29. memorandum. [cited May 26, 2010]; Available from: http://www.fda.gov/ohrms/dockets/ ac/07/briefi ng/2007-4323b1-02-FDA.pdf

122

EEliflif FatmaFatma SSenen BW.inddBW.indd 112222 003-01-113-01-11 15:1715:17 Chapter 3.3

Use of corticosteroids in children and risk of fractures: a nested case-control study in two European countries

E. Fatma Sen, Katia MC Verhamme, Sandra de Bie, Carlo Giaquinto, Adriana Ceci, and Miriam CJM Sturkenboom, on behalf of the TEDDY European Network of Excellence.

EEliflif FatmaFatma SSenen BW.inddBW.indd 112323 003-01-113-01-11 15:1715:17 Abstract

Background: Asthma medicines, including glucocorticosteroids (inhaled [ICS or oral [OCS] are amongst the most frequently used drugs in childhood. Long term use of these medicines has been associated with several side eff ects, including fractures Objective: To estimate the association between use of inhaled and oral steroid and risk of fractures in children Setting: two primary care information databases in the Europe (IPCI from the Netherlands and Pedianet from Italy) Methods: We conducted a population-based case-control study during 1996-2008, nested in a cohort of users of asthma medicines. The risk of fractures with steroids use was calculated for current, past and cumulative use of ICS, OCS and ICS+OCS. Each case was matched to the maximum number of controls from the asthma medicines users cohort. Odds ratios (ORs) were estimated through conditional logistic regression analyses. Results: The adjusted OR for current exposure to ICS+OCS was 0.80, CI [0.6-1.1] and OR 0.97, CI [0.9-1.1] for past use. For cumulative exposure to ICS+OCS the following ORs were observed: OR 0.93, CI [0.8-1.1] for <60 days; OR 1.00, CI [0.8-1.2] for 60-120 days; OR 0.98, CI [0.8-1.3] for 120-180 days; and OR 0.98, CI [0.8-1.2] for >180 days of use. For cumulative exposure to ICS and OCS alone we also did not fi nd an association. Conclusion: We found no increased risk of fractures for children exposed to ICS alone, OCS alone or combined use of ICS+OCS. Our fi ndings, show that ICS, which are the mainstay of pediatric asthma treatment, can be used in children without fearing for fractures.

124

EEliflif FatmaFatma SSenen BW.inddBW.indd 112424 003-01-113-01-11 15:1715:17 Chapter 3.3: Use of corticosteroids in children and risk of fractures

Introduction

Asthma is defi ned as a chronic infl ammatory disorder of the airways and is associated with airway hyperresponsiveness that leads to recurrent episodes of wheezing, breathlessness, chest tightness, and coughing [1]. In children, asthma is the most common chronic disease and the leading cause of childhood morbidity from chronic disease as measured by school absences, emergency department visits and hospitalizations [2]. According to the Global Initiative for Asthma (GINA), inhaled (gluco)corticosteroids (ICS) are the fi rst choice therapy in the long-term management of asthma in children [1]. Besides ICS, oral (gluco)corticosteroids (OCS) are recommended for the treatment of very severe persistent asthma and exacerbations. As a result, ICS and OCS are widely used in asthmatic children. Prevalence rates of use ranges between 1 and 50 percent per year for ICS and between 1 and 40 percent for OCS, depending on country, age range and underlying condition [3]. Long term use of ICS in children has been associated with major safety concerns such as adrenal suppression and growth retardation [1]. The association between the use of ICS and the risk of fractures remains controversial and very few studies have been conducted in children. This potential association is diffi cult to investigate by means of randomized controlled trials (RCTs) as the follow-up in these trials is often too short and the sample size too small.. To our knowledge, there are 2 case-control studies (both in the UK conducted with the general practice research database) evaluating the association between ICS use and risk of fractures in children, and one cross-sectional study [4-6]. In these studies, use of ICS was not associated with an increased risk of fractures. To our knowledge, there are no studies assessing the association between corticosteroids use and risk of fracture in children using combined data other European countries, where the patterns of use may diff er, for example Italy where use of steroids, especially oral, is relatively high. To quantify the risk of fractures associated with the use of corticosteroids in children, we conducted a population based nested case-control study using two primary care databases: the Dutch Integrated Primary Care Information (IPCI) and the Italian Pedianet database, including children 0-18 years of age.

125

EEliflif FatmaFatma SSenen BW.inddBW.indd 112525 003-01-113-01-11 15:1715:17 Methods

Setting Data were obtained from general practice medical record databases in two countries and were extracted and elaborated according to a common protocol. The databases comprised the Pedianet database in Italy [7] and the Integrated Primary Care Information (IPCI) database in the Netherlands [8]. These databases include complete automated medical records of primary care physicians (general practitioner in the Netherlands and family paediatrician in Italy). As in Italy children are registered with a family paediatrician up to the age of 14, the Pedianet database contains data for children aged 0-14 years. The patient population in each database is representative of the respective Dutch and Italian population regarding age and gender. The electronic records contain anonymous and coded information on patient demographics, symptoms and diagnoses, referrals, clinical fi ndings, laboratory assessments, drug prescriptions, and hospitalisations. Details of these databases with regard to their pediatric data are described elsewhere [9].

Source population The source population consisted of all patients registered in the IPCI and Pedianet database and with at least 365 days of valid history except for newborns who did not require a previous medical history. This meant that the practice had been contributing data to the database for at least 12 months and that the patient had been registered with the primary care physician for at least 12 months. A one year pre-enrollment period was required (except for newborns) in order to be able to characterize the patients in terms of comorbidity and exposure. The study period started on January 1996 and ended in June 2008.

Study cohort The study cohort included all patients who received a prescription for an asthma drug (ATC code R03) before or at the age of 18 years. All patients with a prescription for an asthma drug were followed from cohort entry until the occurrence of a fracture, death or end of the study period, whichever came fi rst.

Cases defi nition and nested case-control analysis Case patients were all patient with a fi rst ascertained bone fracture. We used a 2-step proce- dure for case ascertainment. First, we identifi ed all the potential cases of a fracture through an automatic electronic search on coded diagnosis ( ICPC codes L72-76, ICD-9 codes 800-829) in combination with a free text search in the two databases. Second, all hits of fractures were manually reviewed by two medically trained experts. Only fractures that were confi rmed by an X-ray or medical specialist were included in the analysis. The index date was the date of the fi rst

126

EEliflif FatmaFatma SSenen BW.inddBW.indd 112626 003-01-113-01-11 15:1715:17 Chapter 3.3: Use of corticosteroids in children and risk of fractures

X-ray or specialist confi rmed fracture after study entry. Clavicle fractures occurring during birth were excluded from the analysis. To each case set we matched a maximum number of controls from the asthma drugs users cohort on index date, gender, year of birth and general practice. Due to this incidence density sampling approach, controls could be re-sampled at diff erent moments in time and their con- tribution should be considered in person-time (moments) rather than by number of subjects.

Defi nition of drug exposure To estimate the association between use of inhaled or oral corticosteroids and fracture, we created exposure categories based on timing and duration of use. We obtained data on corticosteroid use from prescription fi les and calculated duration of exposure on the basis of the prescribed number of units and the dosing regimen. We grouped exposure to steroids in the following categories: exposure to inhaled corticosteroids (ICS, this included prescriptions for ICS only as well as fi xed combinations of ICS+short-acting beta2-agonists), exposure to oral corticosteroids (OCS), and exposure to a combination of inhaled and oral corticosteroids (ICS+OCS). We defi ned drug exposure as ‘current’ if the prescription covered the index date or if the last intake occurred within 30 days prior to the index date and as ‘past’ if the child had received a prescription anytime before index date and last intake ended more that 30 days prior to the index date. To analyse the eff ect of cumulative exposure to corticosteroids and risk of fractures, we used total duration of exposure. Total duration of exposure was defi ned as the total number of days a child was exposed to corticosteroids from start of follow-up until the index date (e.g. if a child received two presciptions prior to the index date namely one for 30 days and one for 60 days, the total days of exposure will be 90 days). If the indexdate fell between the start and stopdate of the prescription, only the number of exposed days prior to the index date were taken into account. The eff ect of cumulative exposure was investigated for ICS+OCS (summing days of exposure for both types), ICS only and OCS only.

Covariates Age, gender, a previous history of trauma, diabetes mellitus, epilepsy, history of biliary tract disease, hyperthyroidism and hypothyroidism were considered as potential confounders. To control for confounding by indication, we also considered other classes of asthma treatment namely inhaled short-acting and long-acting beta2-agonists (ATC R03AC), inhaled anticholinergics (ATC R03BB), leukotriene receptor-antagonists (ATC R03DC), anti-allergics (ATC R03BC) and xanthines (ATC R03DA) Current exposure to the following concomitant drug were also taken into account: drugs for gastrointestinal disorders (ATC A03), drugs for acid related disorders (ATC A02), drugs used in diabetes (ATC A10), psychotropics+antidepressants (ATC N05), cardiovascular drugs (ATC C), sex

127

EEliflif FatmaFatma SSenen BW.inddBW.indd 112727 003-01-113-01-11 15:1715:17 hormones (ATC G03), anti-epileptics (ATC N03), anti-infl ammatory and anti-rheumatic drugs (ATC M01).

Statistical analysis We compared characteristics of cases and controls by using conditional logistic regression. Covariates that were associated (P < 0.05) with fractures were later added to the fi nal adjusted model. We used conditional logistic regression analysis to assess the matched unadjusted and adjusted estimates of risk for the association between exposure to steroids and the occurrence of a fracture. We estimated odds ratios (ORs) and 95% confi dence intervals (95% CI) for current and total exposure to steroids compared to no exposure. We performed stratifi ed analyses by gender, to observe whether gender modifi ed the risk of steroids on fractures. We conducted all conditional logistic regression analyses in SPSS/PC, version 13 (SPSS, Chicago, Illinois). The level of signifi cance for all statistical tests was set at P-value less than 0.05.

Results

The study cohort comprised of 75,385 children, being 18 years or younger at cohort entry, which was a prescription for an asthma drug during the study period. Within this cohort, we identifi ed 2,548 cases of incident fractures. The fracture site in IPCI was as follows: 395 (40.7%) had a fracture of the arm, 215 (22.2%) had a fracture of the hand/fi ngers/wrist, 145 (14.9%) had a fracture of the foot/toe/ankle, 85 (8.8%) had a fracture of the leg and 122 (12.6%) had a fracture at another location. The fracture site was not available for Pedianet. The characteristics of case and control patients with regards to age, gender and covariates are displayed in Table 1. The mean age of the cases was 8.8 years (standard deviation 4.7) and 62.8% were males. Cases had a higher prevalence of trauma and current exposure to drugs for functional gastrointestinal disorders, sex hormones and anti-infl ammatory and anti-rheumatic drugs increased the risk of a fracture. These covariates were adjusted for in the fi nal model. Exposure to other asthma drugs is displayed in Table 2. The most commonly used other asthma drugs were short-acting beta2-agonists and anticholinergics. Current and past use of other classes of asthma drugs were not associated with a higher risk of fractures. Exposure to corticosteroids and the risk of fractures is displayed in Table 3, the adjusted OR for current exposure to ICS+OCS was 0.80, CI [0.6-1.1] and OR 0.97, CI [0.9-1.1] for past use. When considering total days of exposure to ICS+OCS, no association was observed between the cumulative days of exposure and the risk of fractures: OR 0.93, CI [0.8-1.1] for <60 days; OR 1.00, CI [0.8-1.2] for 60-120 days; OR 0.98, CI [0.8-1.3] for 120-180 days; and OR 0.98, CI [0.8-1.2] for

128

EEliflif FatmaFatma SSenen BW.inddBW.indd 112828 003-01-113-01-11 15:1715:17 Chapter 3.3: Use of corticosteroids in children and risk of fractures

Table 1: Risk of fractures according to patient characteristics. * Values are numbers (%) unless stated otherwise Cases Controls Matched OR# Characteristics p-value (n=2,548) (n=32,778) (95% CI) Age (mean,SD) 8.8 (4.7) 6.9 (3.6) Male 1,600 (62.8) 21,270 (64.9) Comorbidity Trauma 168 (6.6) 203 (0.6) 11.19 (8.85-14.15) <0.0001 Current use of other drugs (use in month before index date) Drugs for functional gastrointestinal disorders (ATC 6 (0.2) 117 (0.4) 1.12 (1.03-1.23) 0.01 A03) Sex hormones (ATC G03) 47 (1.8) 111 (0.3) 1.43 (1.06-1.93) 0.02 Anti-infl ammatory and anti- 24 (0.9) 269 (0.8) 1.09 (1.01-1.18) 0.03 rheumatic drugs (ATC M01) * Only variables with cell count ≥ 5 and p-value <0.05 are shown # = matched on gender, age, index date and general practice

Table 2: Risk of fractures according to use of respiratory drugs. Values are numbers (%) unless stated otherwise Cases Controls Matched OR# p-value (n=2,548) (n=32,778) (95% CI) Inhaled short-acting beta-agonists Never used 1,151 (45.2) 17,940 (54.7) 1 (Reference) - Current use 102 (4.0) 446 (1.4) 1.05 (0.80-1.36) 0.74 Past use 1,295 (50.8) 14,392 (43.9) 1.05 (0.95-1.16) 0.31 Inhaled long-acting beta-agonists Never used 2,478 (97.3) 32,457 (99.0) 1 (Reference) - Current use 6 (0.2) 23 (0.1) 0.68 (0.22-2.06) 0.49 Past use 64 (2.5) 298 (0.9) 1.14 (0.83-1.56) 0.42 Anticholinergics Never used 2,437 (95.6) 31,834 (97.1) 1 (Reference) - Current use 8 (0.3) 22 (0.1) 1.39 (0.51-3.81) 0.52 Past use 103 (4.0) 922 (2.8) 1.00 (0.77-1.31) 0.99 Anti-allergics Never used 2,474 (97.1) 32,066 (97.8) 1 (Reference) - Current use 4 (0.2) 24 (0.1) 1.03 (0.29-3.62) 0.96 Past use 70 (2.7) 688 (2.1) 0.85 (0.64-1.14) 0.28 Leukotriene-receptor antagonists Never used 2,471 (97.0) 31,652 (96.6) 1 (Reference) - Current use 19 (0.7) 206 (0.6) 1.48 (0.89-2.47) 0.13 Past use 58 (2.3) 920 (2.8) 1.15 (0.87-1.53) 0.34 Xanthines Never used 2,534 (99.5) 32,392 (98.8) 1 (Reference) - Current use 0 1 (0.0) NA - Past use 14 (0.5) 385 (1.2 1.03 (0.57-1.86) 0.93 NA = not applicable, as no exposed cases # = matched on gender, age, index date and general practice

129

EEliflif FatmaFatma SSenen BW.inddBW.indd 112929 003-01-113-01-11 15:1715:17 Table 3: Exposure to steroids (inhaled or oral) and the risk of fracture Matched OR#, Matched OR#, Cases Controls Crude adjusted* p-value (n=2,548) (n=32,778) (95% CI) (95% CI) Never used steroids 526 (20.6) 4,022 (12.3) 1 (Reference) 1 (Reference) - (inhaled+oral) Use of inhaled+oral steroids Current 92 (3.6) 573 (1.7) 0.81 (0.62-1.06) 0.80 (0.61-1.05) 0.10 Past 1,930 (75.7) 28,183 (86) 0.97 (0.86-1.10) 0.97 (0.86-1.10) 0.62 Total days of use Inhaled+oral steroids <60 days 1,381 (54.2) 24,378 (74.4) 0.92 (0.80-1.06) 0.93 (0.80-1.07) 0.28 60-120 days 245 (9.6) 2,180 (6.7) 1.00 (0.84-1.21) 1.00 (0.83-1.20) 0.98 120-180 days 99 (3.9) 704 (2.1) 1.01 (0.78-1.31) 0.98 (0.76-1.27) 0.89 >180 days 297 (11.7) 1,494 (4.6) 0.99 (0.83-1.18) 0.98 (0.82-1.17) 0.80 Inhaled steroids <60 days 1,322 (51.9) 23,322 (71.2) 0.93 (0.80-1.07) 0.93 (0.81-1.08) 0.33 60-120 days 231 (9.1) 1,863 (5.7) 0.99 (0.82-1.20) 0.99 (0.82-1.19) 0.88 120-180 days 92 (3.6) 664 (2.0) 0.96 (0.74-1.25) 0.93 (0.71-1.22) 0.61 >180 days 295 (11.6) 1,429 (4.4) 1.00 (0.83-1.19) 0.98 (0.82-1.18) 0.85 Oral steroids <7 days 362 (14.2) 6,897 (21.0) 0.96 (0.81-1.13) 0.95 (0.80-1.13) 0.56 7-14 days 120 (4.7) 2,089 (6.4) 1.01 (0.80-1.27) 0.96 (0.76-1.21) 0.73 14-30 days 52 (2.0) 1,204 (3.7) 0.92 (0.67-1.28) 0.86 (0.62-1.20) 0.38 >30 days 26 (1.0) 464 (1.4) 1.18 (0.73-1.89) 1.17 (0.73-1.89) 0.52 # = matched on gender, age, index date and general practice * = adjusted for use of drug for functional gastrointestinal disorders, sex hormones, anti-infl ammatory and anti-rheumatic drugs and trauma

>180 days of use. Although almost children had long term exposure to ICS (50% of the children had more than 2 months of ICS exposure, and 11% more than 6 months), no association was observed for cumulative duration of ICS exposure. Cumulative exposure of OCS alone was also not associated (although few children had long term exposure to OCS). In the stratifi ed analysis by gender, the results were similar to the overall results (Table 4). No association was observed for exposure to steroids (inhaled or oral) and risk of fractures in males and females.

Discussion In this population based nested case-control study in an Italian and Dutch cohort of pediatric asthma drug users, we found no increased risk of fractures for children exposed to ICS alone, OCS alone or combined use of ICS+OCS. Exposure to OCS is a known risk factor for developing fractures, both in children and adults [10-12]. However, only few studies have investigated the association between ICS use and the risk of fractures in children. To our knowledge, there are only 3 observational studies that have

130

EEliflif FatmaFatma SSenen BW.inddBW.indd 113030 003-01-113-01-11 15:1715:17 Chapter 3.3: Use of corticosteroids in children and risk of fractures

Table 4: Exposure to steroids (inhaled or oral) and the risk of fracture stratifi ed by gender Males Females Matched Matched Cases Controls OR#, Cases Controls OR#, p-value p-value (n=1,600) (n=21,270) adjusted* (n=948) (n=11,508) adjusted* (95% CI) (95% CI) Never used 2,540 1 207 1,482 1 steroids 319 (19.9) - - (11.9) (Reference) (21.8) (12.9) (Reference) (inhaled+oral) Use of inhaled+oral steroids 0.83 0.73 Current 63 (3.9) 416 (2.0) 0.25 29 (3.1) 157 (1.4) 0.22 (0.59-1.15) (0.44-1.20) 1,218 18,314 0.96 712 9,869 0.98 Past 0.62 0.88 (76.1) (86.1) (0.82-1.13) (75.1) (85.8) (0.80-1.21) Total days of use Inhaled+oral steroids 15,679 0.91 518 8,699 0.95 <60 days 863 (53.9) 0.32 0.63 (73.7) (0.76-1.09) (54.6) (75.6) (0.75-1.19) 0.90 100 1.19 60-120 days 145 (9.1) 1,495 (7.0) 0.39 685 (6.0) 0.26 (0.71-1.14) (10.5) (0.75-1.19) 1.03 0.87 120-180 days 70 (4.4) 519 (2.4) 0.88 29 (3.1) 185 (1.6) 0.57 (0.75-1.40) (0.54-1.41) 1.02 0.89 >180 days 203 (12.7) 1,037 (4.9) 0.88 94 (9.9) 457 (4.0) 0.47 (0.81-1.27) (0.66-1.22) Inhaled steroids 15,030 0.93 492 8,292 0.94 <60 days 830 (51.9) 0.40 0.64 (70.7) (0.78-1.11) (51.9) (72.1) (0.75-1.20) 0.92 1.11 60-120 days 140 (8.8) 1,279 (6.0) 0.51 91 (9.6) 584 (5.1) 0.52 (0.73-1.17) (0.81-1.51) 0.91 0.89 120-180 days 63 (3.9) 488 (2.3) 0.71 29 (3.1) 176 (1.5) 0.65 (0.68-1.30) (0.55-1.45) 1.02 0.90 >180 days 201 (12.6) 991 (4.7) 0.84 94 (9.9) 438 (3.8) 0.50 (0.82-1.28) (0.66-1.23) Oral steroids 4,644 0.92 129 2,253 1.02 <7 days 233 (14.6) 0.41 0.90 (21.8) (0.74-1.13) (13.6) (19.6) (0.77-1.35) 0.93 1.03 7-14 days 79 (4.9) 1,390 (6.5) 0.61 41 (4.3) 699 (6.1) 0.89 (0.69-1.24) (0.69-1.52) 0.83 0.92 14-30 days 36 (2.3) 835 (3.9) 0.36 16 (1.6) 369 (3.2) 0.77 (0.56-1.24) (0.51-1.64) 1.14 1.25 >30 days 19 (1.2) 326 (1.5) 0.66 7 (0.7) 138 (1.2) 0.63 (0.65-2.00) (0.50-3.12) # = matched on gender, age, index date and general practice * = adjusted for use of drug for functional gastrointestinal disorders, sex hormones, anti-infl ammatory and anti-rheumatic drugs and trauma

analysed this association. Two of these studies used the General Practice Research Database (GPRD) from the United Kingdom to perform nested case-control studies [4, 5]. In the fi rst study, Schlienger et.al found no increased risk of fractures in children currently exposed to ICS and the risk did not increase for higher number of ICS prescriptions [4]. In the second case-control

131

EEliflif FatmaFatma SSenen BW.inddBW.indd 113131 003-01-113-01-11 15:1715:17 study in the same database, Van Staa et. al found a dose-response relationship namely that, children exposed to higher doses of ICS had an increased risk of fractures [5]. This association disappeared however after adjustment for indicators of asthma severity. In a cross-sectional study in Australia, Ma et al. studied risk factors for prevalent fractures in prepubertal children. In this study, exposure to ICS in the year before occurrence of a fracture was not associated with an increased risk of fractures. Van Staa et al. also studied the association between OCS use and fractures in the GPRD in the UK and found that the risk was increased in children receiving more than 3 prescriptions [10]. In contrast to the fi ndings from Van Staa, we did not fi nd an increased risk associated with exposure to OCS. However, this can probably be explained by the fact that, in our study; the majority of the children used OCS for less than 30 days. Indeed, only 1% of the children (both cases and controls) used OCS longer than one month, so our exposure to long term OCS may be too low to fi nd an association. It does show however, that under the general circumstances of use in these two countries, which have quite diff erent user patterns, the attributable risk of fractures due to use of ICS and/or OCS is very low. Similar to what has been published in literature; we also found that occurrence of fractures was higher in boys than in girls [13-15]. Our results are also in line with other studies regarding the distribution of fracture sites, showing that fractures of the arm and carpal region are the most common [14, 15].

Strengths and limitations of our study This study has a number of limitations. Due to the nature of the databases, we did not have information on all potential confounders such as the nutritional status of the children, body mass index, bone mineral density, physical activity and smoking status. The inability to completely adjust for confounding normally results in an overestimation of the risk estimate. However, as we did not observe an increased risk of fractures in the matched nor in the fully adjusted analysis, we believe that the issue of remaining confounding is probably limited. Selection bias was unlikely, as all data was obtained from prospectively collected medical records that are maintained for patient care purposes. Information bias by misclassifi cation of the outcome is also unlikely, because all fracture cases were retrieved from medical records and reviewed by two medical experts who were blinded to the exposure. Misclassifi cation of exposure may have occurred, because we used outpatient prescription data and had no information on actual fi lling of the medications nor on drug intake. Based on the way the data were collected, we believe that this misclassifi cation will probably be non-diff erential between cases and controls and therefore the actual risk may have been underestimated. Since we used primary care data we may have missed some specialist prescriptions in the Netherlands. However, this will minimally infl uence our results since asthma is a condition often dealt with in primary care, and asthma drugs originally prescribed by a specialist are often continued or prescribed by general practitioners [16].

132

EEliflif FatmaFatma SSenen BW.inddBW.indd 113232 003-01-113-01-11 15:1715:17 Chapter 3.3: Use of corticosteroids in children and risk of fractures

A major strength of our study is the fact that we capture a large number of children in two diff erent databases across Europe, resulting in a high generalisability of the study population. Furthermore, we were able to study the eff ect of total duration of exposure to corticossteroids and the risk of fractures, which has never been studied before. Our fi ndings, as well as other studies, show that ICS, which are the mainstay of pediatric asthma treatment, can be used in children without fearing for fractures. Also OCS do not seem the increase the risk of fractures in children under the given circumstances of use, we can however not exclude that chronic use will not increase the risk since we had few persons exposed to long term use of OCS.

133

EEliflif FatmaFatma SSenen BW.inddBW.indd 113333 003-01-113-01-11 15:1715:17 Acknowledgements: We thank all of the physicians contributing data to the PEDIANET and IPCI databases.

Funding: Funded by the European Community’s 6th Framework Programme. Project number LSHB-CT-2005-005216: TEDDY: Task force in Europe for Drug Development for the Young.

134

EEliflif FatmaFatma SSenen BW.inddBW.indd 113434 003-01-113-01-11 15:1715:17 Chapter 3.3: Use of corticosteroids in children and risk of fractures

References

1 The Global Strategy for Asthma Management and Prevention, Global Initiative for Asthma (GINA) 2007. [cited; Available from: http://www.ginasthma.org 2 Masoli M, Fabian D, Holt S, Beasley R. The global burden of asthma: executive summary of the GINA Dissemination Committee report. Allergy. 2004 May;59(5):469-78. 3 Sen EF, Verhamme KM, Neubert A, Hsia Y, Murray M, Felisi M, et al. Assessment of Pediatric asthma drug use in three European countries; a TEDDY study. European journal of pediatrics. Sep 2. 4 Schlienger RG, Jick SS, Meier CR. Inhaled corticosteroids and the risk of fractures in children and adolescents. Pediatrics. 2004 Aug;114(2):469-73. 5 van Staa TP, Bishop N, Leufkens HG, Cooper C. Are inhaled corticosteroids associated with an increased risk of fracture in children? Osteoporos Int. 2004 Oct;15(10):785-91. 6 Ma DQ, Jones G. Clinical risk factors but not bone density are associated with prevalent fractures in prepubertal children. Journal of paediatrics and child health. 2002 Oct;38(5):497-500. 7 Menniti-Ippolito G, Raschetti R, Da Cas R, Giaquinto C, Cantarutti L. Active monitoring of adverse drug reactions in children. Italian Paediatric Pharmacosurveillance Multicenter Group. Lancet. 2000;355:1613-4. 8 Vlug AE, van der Lei J, Mosseveld BM, van Wijk MA, van der Linden PD, Sturkenboom MC, et al. Post- marketing surveillance based on electronic patient records: the IPCI project. Methods of information in medicine. 1999 Dec;38(4-5):339-44. 9 Sturkenboom MC, Verhamme KM, Nicolosi A, Murray ML, Neubert A, Caudri D, et al. Drug use in children: cohort study in three European countries. BMJ (Clinical research ed. 2008;337:a2245. 10 van Staa TP, Cooper C, Leufkens HG, Bishop N. Children and the risk of fractures caused by oral corticosteroids. J Bone Miner Res. 2003 May;18(5):913-8. 11 Van Staa TP, Leufkens HG, Abenhaim L, Zhang B, Cooper C. Use of oral corticosteroids and risk of fractures. J Bone Miner Res. 2000 Jun;15(6):993-1000. 12 Vestergaard P, Rejnmark L, Mosekilde L. Fracture risk associated with systemic and topical corticosteroids. Journal of internal medicine. 2005 Apr;257(4):374-84. 13 Landin LA. Fracture patterns in children. Analysis of 8,682 fractures with special reference to incidence, etiology and secular changes in a Swedish urban population 1950-1979. Acta orthopaedica Scandinavica. 1983;202:1-109. 14 Lyons RA, Delahunty AM, Kraus D, Heaven M, McCabe M, Allen H, et al. Children’s fractures: a population based study. Inj Prev. 1999 Jun;5(2):129-32. 15 Cooper C, Dennison EM, Leufkens HG, Bishop N, van Staa TP. Epidemiology of childhood fractures in Britain: a study using the general practice research database. J Bone Miner Res. 2004 Dec;19(12):1976- 81. 16 van der Lei J, Duisterhout J, Westerhof H, van der Does E, Cromme P, Boon W, et al. The introduction of computer-based patient records in the Netherlands. Ann Intern Med. 1993;119(10):1036-41.

135

EEliflif FatmaFatma SSenen BW.inddBW.indd 113535 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 113636 003-01-113-01-11 15:1715:17 Chapter 3.4

Steroid induced pancreatitis in children: results from a population-based case- control study

E. Fatma Sen, Katia MC Verhamme, Carlo Giaquinto, Adriana Ceci, and Miriam CJM Sturkenboom, on behalf of the TEDDY European Network of Excellence.

EEliflif FatmaFatma SSenen BW.inddBW.indd 113737 003-01-113-01-11 15:1715:17 Abstract

Background: Pancreatitis is a relatively rare condition in children and has considerable morbidity and mortality. Althoud drug-induced pancreatitis is rare, a variety of drugs have been associated to pancreatitis (in children and adults), including steroids. So far, no studies have been performed to assess the association between exposure to steroids and pancreatitis in children. Objective: To estimate the association between use of inhaled and oral steroid and risk of pancreatitis in children aged 0-18 years Setting: the Dutch PHARMO Record Linkage System Methods: We conducted a population-based case-control study during 1998-2008. The risk of pancreatitis with steroids use was calculated for current, past and cumulative exposure to inhaled glucocortocosteroids (ICS) and oral glucocorticosteroids (OCS). Up to 20 controls per case were selected, matched on year of birth, gender, and calendar time. Odds ratios (ORs) were estimated through conditional logistic regression analyses. Results: 60 cases of pancreatitis (54 acute and 6 chronic) were retrieved from the PHARMO database. Current use of OCS was associated with a higher risk of pancreatitis (OR 21.40, CI [7.1-64.6]). We also found an increased risk for cumulative exposure of >7 days to OCS. Exposure to inhaled steroids was not associated with pancreatitis. In a sensitivity analysis, performed in patients with acute pancreatitis, exposure to OCS and acute pancreatitis remained highly increased: OR 24.0, CI [7.8-73]. Conclusion: We found that the risk of pancreatitis was approximately 20-fold higher in children who were using oral steroids. The use of inhaled steroids was not associated with an increased risk, neither with increased duration of use.

138

EEliflif FatmaFatma SSenen BW.inddBW.indd 113838 003-01-113-01-11 15:1715:17 Chapter 3.4: Steroid induced pancreatitis in children

Introduction

Pancreatitis is a relatively rare condition in children and has considerable morbidity and mortality [1]. Pancreatitis mainly occurs in adult life and the mean age of a fi rst pancreatitis attack has been shown to be in the 6th decade of life [2]. The overall incidence of acute pancreatitis varies between 4 and 45/100,000, depending on the country being studied. In adults, the most common causes of acute pancreatitis are gallstones and alcohol abuse [2, 3]. In children, acute pancreatitis has diff erent aetiologies. Reported causes in children include trauma, infection, systemic diseases, toxins, structural anomalies of the pancreaticobiliary tract and idiopathic [4-6]. Drug-induced pancreatitis is considered to be a rare cause of pancreatitis in both children and adults. The available data suggest that between 0.1 and 2% of all acute pancreatitis cases may be drug-induced [7-9]. A variety of drugs have been associated to pancreatitis through spontaneous reports in adults, including glucocorticosteroids [7, 10, 11]. These drugs have also been associated with pancreatitis in children [5, 6, 12]. The pathogenesis of steroid-induced pancreatitis remains to be elucidated. Obstruction of the outfl ow of pancreatic secretions has been hypothesized as a possible underlying mechanism [13-15]. In contrast, animal studies and a study in humans have reported on the protective eff ect of steroids on pancreatitis. Experimen- tal studies in rats have shown a protective eff ect of dexamethason and methyprednisolone on the development and severity of pancreatitis [16, 17]. The real association between exposure to steroids and risk of pancreatitis remains to be elucidated and more studies are needed. All of the existing data on the association between glucocorticosteroid use and pancreatitis in children has been derived from case reports and case series. Although case reports are a very effi cient tool for signal generation, more analytical and epidemiological approaches are necessary to actually quantify the risk. To provide the relative risk of pancreatitis in children using oral or inhales corticosteroids, we conducted a population-based case-control study. To our knowledge, this is the fi rst study in children analyzing the association between use of glucocorticosteroids and the risk of pancreatitis.

Methods

Source population Data for this case control study were obtained from the Dutch PHARMO Record Linkage System (RLS). PHARMO RLS (http://www.pharmo.nl) includes the pharmacy dispensing histories of more than three million community-dwelling residents in well defi ned areas in the Netherlands, and these pharmacy records are linked to nationwide hospital discharge records. Pharmacy data include information about the drug dispensed, the date of dispensing, the amount dispensed, the prescribed dosage regimen and the estimated duration of use. Drugs are coded according to the Anatomical Therapeutic Chemical (ATC) classifi cation system [11]. The Hospital Discharge

139

EEliflif FatmaFatma SSenen BW.inddBW.indd 113939 003-01-113-01-11 15:1715:17 Register comprises hospitalizations in the Netherlands, including detailed information on the primary and secondary discharge diagnoses; diagnostic, surgical and treatment procedures and dates of hospital admission and discharge. All diagnoses are coded according to the Inter- national Classifi cation of Diseases, 9th edition (ICD-9-CM). Validation studies have shown that the PHARMO RLS has a high level of data completeness and validity [18, 19].

Case and control ascertainments. Cases were defi ned as patients 0- ≤18 years old with a fi rst discharge diagnosis of pancreatitis (ICD-9 code 577x) in the period between 1 January 1998 (start of follow-up) until 31 December 2008 (end of follow-up). Pancreatitis was classifi ed as acute (ICD-9 577.0) or chronic (ICD-9 577.1). The index date was the date of hospital admission. For each case, up to 20 controls without a history of pancreatitis during the study period were matched on year of birth, gender, and calendar time (the index date of the corresponding case). Both cases and controls were eligible for inclusion if they had a minimum period of 365 days of history in the PHARMO RLS prior to the index date.

Exposure assessment Although various drugs have been associated with pancreatitis the main exposure of interest in this study were inhaled (ATC R03BA) and oral glucocorticosteroids (ATC H02AB). Data on exposure were obtained from the dispensing records and categorized by timing and duration of use. We defi ned drug use as ‘current’ if the prescribed duration of the drug covered the index date or if the end of it was less than 30 days prior to the index date. Drug exposure was classifi ed as ‘past’ if the child had received a prescription anytime before the index date but the end of it was more than 30 days prior to the index date. To analyse the eff ect of the cumulative use of glucorticosteroids on the risk of pancreatitis, we calculated the total cumulative duration of exposure from start of follow-up until the index date (e.g. if a child received 2 prescriptions prior to the index date namely one for 4 days and one for 6 days, the total days of use was 10). The total days of use of oral steroids was categorized into the following classes: <7 days, ≥ 7-30 days and ≥ 30 days of use. For inhaled steroids we created two categories for total days of use, namely <1 year and ≥ 1 year. If the indexdate fell between the start and stopdate of the prescription, only the number of days up to the index date was taken into account.

Covariates As covariates, we considered co-morbidities or use of drugs that have associated with pancreatitis in literature [7, 20]. The following factors were considered: gender and year of birth (matching variables), a previous hospitalization for HIV-infection, gallbladder disorders, chole- lithiasis, trauma, cystic fi brosis, lipid disorders, disorders of calcium metabolism, infl ammatory bowel disease, malignancies, elevated levels of lipase/amylase and viral diseases.

140

EEliflif FatmaFatma SSenen BW.inddBW.indd 114040 003-01-113-01-11 15:1715:17 Chapter 3.4: Steroid induced pancreatitis in children

Current use of the following concomitant drugs were also taken into account: acetaminophen, azathioprine, cimetidine, enalapril, erythromycin, estrogens, furosemide, hydrochlorothiazide, metronidazol, opiates, simvastatin, cotrimoxazol, tetracycline, valproic acid, carbamazepine, doxycycline, famotidine, mesalazine, rifampin, NSAID’s, lisinopril, ramipril, losarten, methyldopa, chlorothiazide, amiodarone, pravastatin, fl uvastatin, dapsone, nitrofurantoin, clarithromycin, isoniazid, nelfi navir, lamivudine, didanosene, interferon, ribavirin, pentamidine, sodium stibo- gluconate, meglumine antimonite, omeprazole, mercaptopurine, olsalazine, sulphasalazine, clozapine, clomifene, tamoxifen, thiamazole, carbimazole, ifosfamide, cytarabine, asparaginase, isotretinoin, propofol, cyclopenthiazide, phenformin, cisplatin, octreotide, alendronate, capto- pril, ciprofi brate and maprotiline.

Statistical analysis We used conditional logistic regression analysis to assess the matched unadjusted estimates of risk for the association between exposure to inhaled and oral steroids and the occurrence of a pancreatitis. Adjusted analysis was performed only if more than 5 cases were exposed. For model construction, we fi rst included all covariates that were univariately associated with the outcome (p<0.10). Each risk factor that changed the OR of exposure to glucocorticosteroids by more than 5% was included in the fi nal adjusted model. We estimated odds ratios (ORs) and 95% confi dence intervals (95% CI) for current and total use of steroids compared to no use. Sensitivity analyses were conducted in which we excluded patients with chronic pancreatitis. All statistical analysis were conducted using SPSS/PC, version 13 (SPSS, Chicago, Illinois). The level of signifi cance for all statistical tests was set at p-value less than 0.05.

Results

In the PHARMO RLS source population of children 0-18 years of age we identifi ed 68 cases with pancreatitic disorders. Eight cases were excluded because they concerned patients with pancreatic cysts. The case-control study population therefore consisted of 60 patients with pancreatitis (acute or chronic), ≤ 18 years with a fi rst hospitalization for pancreatitis and 6,487 age, sex and time matched controls. The age distribution of the pancreatitis cases is presented in Figure 1. The number of cases increased with age. Details on prior hospitalizations and concomitant use of drugs in the cases is presented in Table 1, 54 cases (90%) were hospitalized for acute pancreatitis and 6 for chronic pancreatitis. The median age of the cases was 15.9 (range between 1.2 and 18.9) and the majority of the cases was between 12 and 18 years of age (71.7%), gender was equally distributed. Only few patients had been hospitalized prior to their pancreatitis, the most frequent reasons were: abdominal pain (n=5), colitis/gastro-enteritis (n=4) and infl ammatory bowel disease

141

EEliflif FatmaFatma SSenen BW.inddBW.indd 114141 003-01-113-01-11 15:1715:17 16

4 number pancreatitis of cases 2

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Age at indexdate Figure 1: Distribution of pancreatitis cases by age

(n=3). The average duration between hospitalization for abdominal pain and hospitalization for pancreatitis was 6.3 months (range between 4 and 11 months) The most frequently used drugs in the month before the index date were propulsives (e.g. metoclopramide, cisapride) (n=4), oral contraceptives (n=4), anti-acne preparations (n=3) and azathioprine (n=3), 29 cases (49%) had not any drug dispensed that covered the month before the index date. A hospitalization for infl ammatory bowel disease (OR 41.9, 95% CI [9-188]), gallstones (OR 120.6, 95% CI [11-1368]) and cystic fi brosis (OR 29.5, 95% CI [3-253]) were all associated with pancreatitis although numbers were small (table 2). Also exposure to NSAIDs (OR 4.5, 95% CI [2-13), azathioprine (OR 37.8, 95% CI [9-154), mesalazine (OR 69.4, 95% CI [10-466]), metronidazol (OR 69.4, 95% CI [10-466]), opiates (OR 26.9, 95% CI [3-287]) and omeprazole (OR 11.3, 95% CI [1.4-88]) in the month prior to the index date was associated with pancreatitis (Table 2). Four cases were currently exposed to oral steroids and one to inhaled steroids (table 3). Due to the low number of exposed cases, adjustment for concomitant drug use or a history of prior hospitalizations was not possible. Despite the low power, current use of oral steroids was associated with a higher risk of pancreatitis (OR 21.40, CI[7.1-64.6]). For the total days of oral steroids use we found the following risk estimates: OR 2.37, CI [0.32-17.7] for <7 days; OR 5.90, CI [1.8-19.7] for ≥7-<30 days; and OR 5.34, CI [1.9-15.3] for ≥ 30 days of use (Table 3). Exposure to inhaled steroids was not associated with pancreatitis in any of the exposure categories. Since steroids may be used to treat pancreatitis, we restricted the analysis to acute pancreatitis, the association between the use of oral steroids and acute pancreatitis remained highly increased: OR 24.0, CI [7.8-73].

142

EEliflif FatmaFatma SSenen BW.inddBW.indd 142142 003-01-113-01-11 15:1715:17 Chapter 3.4: Steroid induced pancreatitis in children

Table 1: Characteristics of pancreatitis cases in children (0-18 years) Pancreatitis cases (n=60)

Gender: Male 29 (48.3%) Age (mean,range) 14.0 (1-18) 0 - < 2 years 1 (1.7%) 2 - ≤ 11 years 16 (26.7%) 12 - ≤ 18 years 43 (71.7%) Type of pancreatitis Acute 54 (90%) Chronic 6 (10%)

Prior hospitalizations Abdominal pain 5 Rheumatoid arthritis 1 Colitis/Gastro-enteritis/proctocolitis 4 Mononucleosis infectiosa 1 Infl ammatory bowel disease 3 Obesity 1 Gallstones/cholecystitis 2 Oesophagitis 1 Juvenile Diabetes 2 Paralytic ileus 1 Cystic Fibrosis 1 Pneumonia 1 Hernia inguinalis 1 Viral infection 1 Intestinal disaccharidase defi ciency 1

Concomitants drugs (use in month before index date) No use of any drug 29 chloral hydrate 1 Oral contraceptives 4 Codeine 1 prednisolon 4 ciprofl oxacin 1 propulsives 4 fenoterol+ipratropium 1 anti-acne preparations 3 ferrous fumarate 1 azathioprine 3 fl uticason 1 diclofenac 3 folic acid 1 insulins 2 granisetron 1 mesalazine 2 ibuprofen 1 metronidazole 2 lamotrigine 1 nasal corticosteroids 2 laurilsulfate 1 oxcarbazepine 2 methotrexate 1 proton pump inhibitor 2 methylphenidate 1 short-acting beta-agonists 2 multi-enzymes 1 sodium fl uoride 2 Omeprazole 1 acetazolamide 1 Paracetamol 1 anti-androgens+estrogens 1 Somatropin 1 azithromycin 1 Temazepam 1 bisacodyl 1 Tobramycin 1 budesonide, inhaled 1 Tramadol 1 calcium 1 vitamin D 1 ceftazidime 1 vitamin K 1

143

EEliflif FatmaFatma SSenen BW.inddBW.indd 114343 003-01-113-01-11 15:1715:17 Table 2: Covariates and their independent risk of pancreatitis in univariate analyses# Characteristics Cases n=60 (%) Controls n=6,487 (%) Matched OR (95% CI)* P-value History of hospitalizations Infl ammatory bowel disease 3 (5.0) 6 (0.1) 42 (9-188) <0.001 Gallstones 2 (3.3) 3 (0.04) 121 (11-1368) <0.001 Cystic fi brosis 1 (1.7) 5 (0.1) 29 (3-253) 0.002 Trauma 1 (1.7) 19 (0.3) 6.3 (0.8-49) 0.08 Viral diseases 1 (1.7) 61 (0.9) 2.2 (0.3-17) 0.44

Concomitant drugs (use in month before index date) NSAIDs 4 (6.7) 103 (1.6) 4.5 (2-13) 0.007 Azathioprine 3 (5.0) 8 (0.1) 38 (9-154) <0.001 Mesalazine 2 (3.3) 3 (0.04) 69 (10-466) <0.001 Metronidazol 2 (3.3) 3 (0.04) 69 (10-466) <0.001 Opiates 1 (1.7) 3 (0.04) 26 (3-287) <0.001 Omeprazole 1 (1.7) 12 (0.2) 11 (1.4-88) 0.021 Paracetamol 1 (1.7) 61 (0.9) 2.2 (0.3-17) 0.44 # = only covariates with exposed cases are shown * = matched on age, gender and index date

Discussion

In this population based case-control study, we found that the risk of pancreatitis was approximately 20-fold higher in children who were using oral steroids. The use of inhaled steroids was not associated with an increased risk, neither with increased duration of use. The association between use of oral steroids and pancreatitis has been described in some case series and case reports in children [6, 21, 22]. The fi rst case reports on steroid-induced pancreatitis in children were published in 1957 by Barr et al and Marczynska-Robowska [21, 22], followed by several others in children [6, 23-25]. Although case series and reports provide important information to signal potential adverse drug reactions, they do not provide proof of an association. In some instances, fi ndings from case reports may overestimate the risk of medication-induced pancreatitis This occurred in adults for postmenopausal hormone replacement therapy, COX-2 inhibitors and anti-lipemics where data from population-based epidemiological studies could not confi rm the associations seen in case series [26-28]. However, for oral steroids we consistently fi nd an association in case series and in this population based case control study. The risk factors we observed in children are in line with those published for pancreatitis in adults. A history of hospitalization for infl ammatory bowel disease, gallstones, cystic fi brosis and trauma were all associated with a higher risk of pancreatitis [4, 29, 30]. We also observed a higher risk of pancreatitis for azathioprine, mesalazine, metronidazol, opiates, omeprazole and NSAIDs use ,but the prevalence of use was very low, which is usually so in children. All

144

EEliflif FatmaFatma SSenen BW.inddBW.indd 114444 003-01-113-01-11 15:1715:17 Chapter 3.4: Steroid induced pancreatitis in children

Table 3: Exposure to steroids (inhaled or oral) and the risk of pancreatitis Cases Controls Matched OR# p-value (n=60) (n=6,847) (95% CI) Never used steroids 46 (76.7) 5,470 (84.3) 1 (Reference) - (inhaled+oral)

Oral steroids Current use 4 (6.7) 24 (0.4) 21.4 (7.1-64.6) <0.001 Past use 4 (6.7) 226 (3.5) 2.56 (0.9-7.3) 0.079 Total cumulative duration of oral steroids <7 days 1 (1.7) 65 (1.0) 2.4 (0.3-17) 0.398 ≥7 -<30 days 3 (5.0) 80 (1.2) 5.9 (1.8-19.7) 0.004 ≥30 days 4 (6.7) 100 (1.5) 5.3 (1.9-15.3) 0.002 Use of inhaled steroids Current 2 (3.3) 201 (3.1) 1.3 (0.6-3.2) 0.54 Past 7 (11.7) 707 (10.9) 1.9 (0.6-6.4) 0.30 Total use <1 year 6 (10.0) 641 (9.9) 4.7 (0.6-35.6) 0.13 ≥1 year 3 (5.0) 253 (3.9) 1.3 (0.6-2.9) 0.46

Subanalysis excluding patients with chronic pancreatitis Cases (n=54) Controls (n=5,957) Use of oral steroids Current 4 (7.4) 22 (0.4) 24.0 (7.8-73.3) <0.001 Past 3 (5.6) 207 (3.5) 2.22 (0.7-7.4) 0.193 Total use <7 days 1 (1.9) 60 (1.0) 2.71 (0.36-20.27) 0.331 ≥7 -<30 days 3 (5.6) 74 (1.2) 6.65 (1.97-22.42) 0.002 ≥30 days 3 (5.6) 90 (1.5) 4.87 (1.46-16.25) 0.01 # = matched on gender, age and index date,

of these drugs have been associated with an increased risk of pancreatitis in adults, although protopathic bias may also occur with the proton pump inhibitors [7, 10, 31]. We did not fi nd an increased risk of pancreatitis for use of inhaled steroids, although power was limited and the upper limit of the confi dence interval was 6. If oral sterids are associated with pancreatitis inhaled steroids may do the same since they are also systemically absorbed, especially when inhaled erroneously. We could demonstrate in this study that at least, there is no strong association [32]. In the future larger scale studies are necessary to further assess the association between pancreatitis and inhaled steroids in children, as it is a frequently used drug by children. The pathogenesis of steroid-induced pancreatitis remains to be elucidated. Obstruction of the outfl ow of pancreatic secretions has been hypothesized as a possible underlying mechanism [13-15]. Animal studies and postmortem studies in patients treated with steroids demonstrated an increased incidence of focal pancreatic lesions [15, 33-35]. In contrast, animal studies and a

145

EEliflif FatmaFatma SSenen BW.inddBW.indd 114545 003-01-113-01-11 15:1715:17 study in humans have also reported on a protective eff ect of steroids on pancreatitis [16, 17] A study in humans has demonstrated that the administration of steroids prior to endoscopic retrograde cholangiopancreatography (ERCP) resulted in a decrease in the incidence of post- ERCP pancreatitis [36]. It is clear that more studies have to be performed to understand the mechanism of steroid-induced pancreatitis. As for all observational studies, our study has limitations, the major one is lack of power due to the rarity of the event in children. This underlines the need for combining large linked healthcare databases to study serious but rare drug-safety issues in children adequately. In addition, we may suff er from residual confounding, Body mass index and alcohol intake are important risk factors of pancreatitis in adults. Alcohol consumption, especially in young children may be low, and therefore it may not be a confounding factor in this study. Confounding by indication may also occur: oral steroids might have been initiated to treat pancreatitis, however, the association was most pronounced in persons using steroids for more than 7 days, therefore, this cannot explain the fi ndings. In addition the association remained upon exclusion of 6 cases with chronic hepatitis. Misclassifi cation of the outcome is a potential concern. All potential cases of pancreatitis were selected from The Hospital Discharge Register via an automated search on ICD-9 codes. However, previous studies have shown the validity of The Hospital Admission Register in identifying patients with diseases of interest [37]. Selection bias is unlikely as we conducted a population based case-control study using a large database containing information on more than 3 million inhabitants from the Netherlands. Our study population is representative to the Dutch population so, our data can be extrapolated to the entire pediatric population of the Netherlands. Exposure misclassifi cation is unlikely as the PHARMO database contains pharmacy dispensing data, on prescriptions from both primary and secondary care, ensuring a complete overview of prescription drugs. The fi rst case reports of steroid-induced pancreatitis in children were published in 1957 [21, 22]. Since then, several case series on this topic have been published, but well-designed observational studies were lacking. We have addressed this gap by performing a population- based case-control study, analyzing the association between steroid use in children and risk of pancreatitis. Our results show that current use and ≥7 days of use of oral steroids is associated with an increased risk of pancreatitis in children.

Funding: Funded by the European Community’s 6th Framework Programme. Project number LSHB-CT-2005-005216: TEDDY: Task force in Europe for Drug Development for the Young.

146

EEliflif FatmaFatma SSenen BW.inddBW.indd 114646 003-01-113-01-11 15:1715:17 Chapter 3.4: Steroid induced pancreatitis in children

References

1 Corfi eld AP, Cooper MJ, Williamson RC. Acute pancreatitis: a lethal disease of increasing incidence. Gut. 1985 Jul;26(7):724-9. 2 Yadav D, Lowenfels AB. Trends in the epidemiology of the fi rst attack of acute pancreatitis: a systematic review. Pancreas. 2006 Nov;33(4):323-30. 3 Whitcomb DC. Clinical practice. Acute pancreatitis. The New England journal of medicine. 2006 May 18;354(20):2142-50. 4 Yeung CY, Lee HC, Huang FY, Ho MY, Kao HA, Liang DC, et al. Pancreatitis in children--experience with 43 cases. European journal of pediatrics. 1996 Jun;155(6):458-63. 5 Alvarez Calatayud G, Bermejo F, Morales JL, Claver E, Huber LB, Abunaji J, et al. Acute pancreatitis in childhood. Rev Esp Enferm Dig. 2003 Jan;95(1):40-4, 5-8. 6 Vane DW, Grosfeld JL, West KW, Rescorla FJ. Pancreatic disorders in infancy and childhood: experience with 92 cases. Journal of pediatric surgery. 1989 Aug;24(8):771-6. 7 Balani AR, Grendell JH. Drug-induced pancreatitis : incidence, management and prevention. Drug Saf. 2008;31(10):823-37. 8 Lankisch PG, Droge M, Gottesleben F. Drug induced acute pancreatitis: incidence and severity. Gut. 1995 Oct;37(4):565-7. 9 Andersen V, Sonne J, Andersen M. Spontaneous reports on drug-induced pancreatitis in Denmark from 1968 to 1999. European journal of clinical pharmacology. 2001 Sep;57(6-7):517-21. 10 Trivedi CD, Pitchumoni CS. Drug-induced pancreatitis: an update. Journal of clinical gastroenterology. 2005 Sep;39(8):709-16. 11 Weaver G. Steroid-induced pancreatitis. Gastroenterology. 1982 Mar;82(3):601. 12 Riemenschneider TA, Wilson JF, Vernier RL. Glucocorticoid-induced pancreatitis in children. Pediatrics. 1968 Feb;41(2):428-37. 13 Anderson MC, Bergan JJ. Signifi cance of vascular injury as a factor in the pathogenesis of pancreatitis. Annals of surgery. 1961 Jul;154:58-67. 14 Nelp WB, Banwell JG, Hendrix TR. Pancreatic function and the viscosity of pancreatic juice before and during cortisone administration. Bulletin of the Johns Hopkins Hospital. 1961 Dec;109:292-301. 15 Stumpf HH, Wilens SL, Somoza C. Pancreatic lesions and peripancreatic fat necrosis in cortisone- treated rabbits. Laboratory investigation; a journal of technical methods and pathology. 1956 Mar- Apr;5(2):224-35. 16 Kandil E, Lin YY, Bluth MH, Zhang H, Levi G, Zenilman ME. Dexamethasone mediates protection against acute pancreatitis via upregulation of pancreatitis-associated proteins. World J Gastroenterol. 2006 Nov 14;12(42):6806-11. 17 Takaoka K, Kataoka K, Sakagami J. The eff ect of steroid pulse therapy on the development of acute pancreatitis induced by closed duodenal loop in rats. Journal of gastroenterology. 2002;37(7):537-42. 18 Heerdink ER, Leufkens HG, Herings RM, Ottervanger JP, Stricker BH, Bakker A. NSAIDs associated with increased risk of congestive heart failure in elderly patients taking diuretics. Archives of internal medicine. 1998 May 25;158(10):1108-12. 19 Pouwels S, Lalmohamed A, Souverein P, Cooper C, Veldt BJ, Leufkens HG, et al. Use of proton pump inhibitors and risk of hip/femur fracture: a population-based case-control study. Osteoporos Int. Jun 29. 20 Tonsi AF, Bacchion M, Crippa S, Malleo G, Bassi C. Acute pancreatitis at the beginning of the 21st century: the state of the art. World J Gastroenterol. 2009 Jun 28;15(24):2945-59.

147

EEliflif FatmaFatma SSenen BW.inddBW.indd 114747 003-01-113-01-11 15:1715:17 21 Barr HS, Wolff OH. Pancreatic necrosis in cortisone-treated children. Lancet. 1957 Apr 20;272(6973):812- 5. 22 Marczynska-Robowska M. Pancreatic necrosis in a case of Still’s disease. Lancet. 1957 Apr 20;272(6973):815-6. 23 Goldberg BH, Bergstein JM. Acute respiratory distress in a child after steroid-induced pancreatitis. Pediatrics. 1978 Feb;61(2):317-8. 24 Di Fazano CS, Messica O, Quennesson S, Quennesson ER, Inaoui R, Vergne P, et al. Two new cases of glucocorticoid-induced pancreatitis. Revue du rhumatisme (English ed. 1999 Apr;66(4):235. 25 Akhtar M, Bhakoo ON, Chandra RK. Pancreatitis in a Child with Leukemia on Intensive Steroid Therapy: Report of a Case. Indian journal of pediatrics. 1964 Nov;31:327-9. 26 Floyd A, Pedersen L, Nielsen GL, Thorlacius-Ussing O, Sorensen HT. Risk of acute pancreatitis in users of fi nasteride: a population-based case-control study. Journal of clinical gastroenterology. 2004 Mar;38(3):276-8. 27 Sorensen HT, Jacobsen J, Norgaard M, Pedersen L, Johnsen SP, Baron JA. Newer cyclo-oxygenase-2 selective inhibitors, other non-steroidal anti-infl ammatory drugs and the risk of acute pancreatitis. Alimentary pharmacology & therapeutics. 2006 Jul 1;24(1):111-6. 28 Tetsche MS, Jacobsen J, Norgaard M, Baron JA, Sorensen HT. Postmenopausal hormone replacement therapy and risk of acute pancreatitis: a population-based case-control study. The American journal of gastroenterology. 2007 Feb;102(2):275-8. 29 Hillemeier C, Gryboski JD. Acute pancreatitis in infants and children. The Yale journal of biology and medicine. 1984 Mar-Apr;57(2):149-59. 30 Weizman Z, Durie PR. Acute pancreatitis in childhood. The Journal of pediatrics. 1988 Jul;113(1 Pt 1):24-9. 31 Nitsche CJ, Jamieson N, Lerch MM, Mayerle JV. Drug induced pancreatitis. Best practice & research. Apr;24(2):143-55. 32 Allen DB. Systemic eff ects of inhaled corticosteroids in children. Current opinion in pediatrics. 2004 Aug;16(4):440-4. 33 Bencosme SA, Lazarus SS. Development and regression of cortisone-induced lesions in rabbit pancreas. American journal of clinical pathology. 1956 Oct;26(10):1146-56. 34 Carone FA, Liebow AA. Acute pancreatic lesions in patients treated with ACTH and adrenal corticoids. The New England journal of medicine. 1957 Oct 10;257(15):690-7. 35 Oppenheimer EH, Boitnott JK. Pancreatitis in children following adrenal cortico-steroid therapy. Bul- letin of the Johns Hopkins Hospital. 1960 Dec;107:297-306. 36 Weiner GR, Geenen JE, Hogan WJ, Catalano MF. Use of corticosteroids in the prevention of post-ERCP pancreatitis. Gastrointestinal endoscopy. 1995 Dec;42(6):579-83. 37 Eland IA, Sturkenboom MJ, Wilson JH, Stricker BH. Incidence and mortality of acute pancreatitis between 1985 and 1995. Scandinavian journal of gastroenterology. 2000 Oct;35(10):1110-6.

148

EEliflif FatmaFatma SSenen BW.inddBW.indd 114848 003-01-113-01-11 15:1715:17 Chapter 4

General Discussion

EEliflif FatmaFatma SSenen BW.inddBW.indd 114949 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 115050 003-01-113-01-11 15:1715:17 Chapter 4: General Discussion

In this thesis we present several large scale observational studies on the use and safety of asthma and cough and cold medicines in children. These studies were conducted in the context of the TEDDY Network of Excellence. TEDDY is a 6th Framework Programme funded Network of Excellence aiming to promote the availability of safe and eff ective medicines for children in Europe by integrating existing expertise and good practices, as well as stimulating pediatric drug development. (www.teddyoung.org) The research described in this thesis focuses on the use and the safety of asthma medicines and cough and cold medicines. These medicines belong to the group of respiratory medicines, which have been demonstrated to be amongst the mostly used medicines in children [1]. Asthma is the most common chronic disease of childhood with an estimated prevalence of 10%, and as a result, asthma medicines are used widely in the pediatric population. Despite the widespread use of these medicines by children, very little is know about their long term safety. For this reason, regulatory authorities have implemented guidelines to stimulate research on safe and eff ective drugs in children, such as the need to get a paediatric investigation plan (PID) for each new regulatory submission and the promotion of the Paediatric Use Marketing Authorisation for off -patent drugs. In addition, the Pediatric Working Party (PEG) (under aus- pices of the EMA) has generated, by treatment class, a list of drugs with specifi c needs, in terms of pediatric effi cacy and safety. Among these lists, a list on the “assessment of paediatric needs for asthma and other obstructive chronic lung diseases” has been compiled stating the need to have information on pharmacokinetics, effi cacy and safety data of respiratory drugs [2]. As proof of concept that these needs could be addressed by using observational studies rather than trials we have conducted large observational utilization and safety studies by combining several primary healthcare databases in Europe.

Other drugs; also frequently used in children; and for which safety concerns have been raised are the cough and cold medicines (CCMs) used for the symptomatic treatment of respiratory tract infections. In 2007, the FDA released a recommendation that cough and cold medicines should not be used in children below two years of age because of serious side eff ects such as cardiac arrhythmias, depressed levels of consciousness and even death [3, 4]. To evaluate the eff ect of these recommendations and to complete the available literature on ADRs caused by CCMs, we conducted several studies as part of the TEDDY project.

151

EEliflif FatmaFatma SSenen BW.inddBW.indd 115151 003-01-113-01-11 15:1715:17 Main Findings

Utilization studies

Use of medicines in children In chapter 2.1 and 2.2 we present the studies on the utilization of pediatric medicines in the Netherlands, Italy and United Kingdom. In chapter 2.1 we describe utilization patterns of all drugs and in chapter 2.2 we focus on the respiratory drug use in children. Using a retrospective cohort design, the utilization of medicines is described for 675,868 children for the years 2000-2005. The off -label status was checked for the fi ve drugs with the highest prevalence, by anatomic class, in each country. A drug was considered to be ‘off -label for age’ if the child’s age at the time of use was below the lowest approved age as mentioned in the Summary of Product Characteristics (SPC) of that drug in each of the countries of use [5]. The databases recorded more than fi ve million pediatric prescriptions. The main fi ndings were the following: 1) the prescription rate was highest for children aged less than two years in all three countries (average of 3.3 prescriptions/year); 2) In children of all ages, the most commonly used medicines were anti-infectives (antibiotics), respiratory medicines (asthma medicines) and dermatologicals (topical corticosteroids); 3) in the younger age categories, boys had a higher user prevalence rate but this pattern reversed with adolescence, where females had consistently higher drug use prevalence rates; 4) off -label use was mostly observed for topical, inhaled and systemic corticosteroids, oral contraceptives and topical or systemic antifungals. This was the fi rst study describing pediatric medicines use in diff erent European countries by combining databases using the same methodology. This approach changed the landscape considerably since previous studies were country or region specifi c with a large heterogeneity between studies, making a comparative evaluation often diffi cult or incomplete [6]. Since we used the same methodology across the databases, the diff erences that were observed refl ect real prescribing diff erences. The Pediatric Committee of the European Medicines Agency (EMA) has compiled a pediatric needs list based on expert opinions and consultations. This list provides an overview of drugs for which additional information on use in children is needed. The missing information can go from PK/PD studies in children to the need of long term safety studies. The drugs mentioned on the list of specifi c needs are also the drugs with the highest user prevalence rates (e.g. inhaled and oral corticosteroids, topical and systemic antifungals). Our fi ndings reinforce the importance of the need of additional data on drugs that are frequently used. However, there are some off -label drugs with high prescription rates that were not mentioned on any of the priority lists of the EMA. One example is the use of sex hormones. Our study revealed high prescription rates and mostly off -label use of these drugs. Similar to adults, the use of oral anticonceptives by teenagers has been associated with an increased risk of cardiovascular events (e.g. myocardial infarction, stroke) as well as with an increased risk of venous thrombo-embolism [7-11]. To our knowledge, no RCTs on the safety and effi cacy of sex

152

EEliflif FatmaFatma SSenen BW.inddBW.indd 115252 003-01-113-01-11 15:1715:17 Chapter 4: General Discussion

hormones limited to teenagers only have been conducted. Therefore we would recommend to include sex hormones in the pediatric needs list. This list is used as a tool to enforce pharmaceutical companies to apply for a PUMA (Pediatric Use Marketing Authorizations) for off -patent drugs to be registered for use in children. If authorized, a PUMA will give the pharmaceutical company 10 years of market protection for that specifi c drug. The studies presented in chapter 2.1 and 2.2 show that drug utilization studies are important to get a clear understanding of the user prevalence rates in children and the diff erences by country, gender and age. Information from these studies is important, fi rst of all to prioritize the specifi c needs but also to be incorporated into the PIPs

Use of asthma medicines Despite the widespread pediatric use of asthma medicines and the large number of single center or national studies describing user prevalence rates, there have been very few comparative studies on the extent and patterns of use by region. To address this, we assessed the use of asthma medicines in three European countries (Netherlands, Italy and United Kingdom) during 2000-2005 (chapter 2.3) by using a similar methodology. In addition, we checked the off -label use of these drugs per country. We found that the prevalence of asthma medicines use was highest in young children and decreased with age. The prescription rates were the highest for most asthma medicines in Italy,

followed by the UK and the Netherlands. ß2-mimetics (salbutamol) and inhaled glucocorticoids (fl uticasone and beclomethasone) were the most frequently used asthma medicines in all three countries. Overall, off -label use, defi ned as use below the minimum age as mentioned in the

SPC was low, but most pronounced for short-acting ß2-mimetics in children < 18 months (Italy)

and fi xed combination of ß2-mimetics+anticholinergics in children <6 years (Netherlands). Some of the respiratory drugs are only registered for the treatment of asthma. Off -label use, defi ned as “use for other indications than asthma”, was observed in all 3 countries for several asthma medicines For some medicines, the percentage of off -label use was considerable (e.g. 80% of the children receiving budesonide in Italy, 74% of the children receiving ipratropium in the UK and 72% of the children receiving salbutamol+ipratropium in the Netherlands do not have a diagnosis of asthma, while these drugs are only licensed for use in asthma). This high percentage of off -label use of asthma medicines in children can be explained by the fact that in young children it is diffi cult to make a diagnosis of asthma. This diagnosis is diffi cult because in young children, it is hard to diff erentiate between respiratory symptoms related to asthma or respiratory symptoms related to lower respiratory tract infections. Second, in this younger age category, asthma diagnosis by means of spirometry is not possible.

Use of cough and cold medicines In 2007 the use of CCMs in young children received a lot of attention from regulatory agencies and scientists; and the Food and Drug Administration (FDA) in the United States issued a

153

EEliflif FatmaFatma SSenen BW.inddBW.indd 115353 003-01-113-01-11 15:1715:17 recommendation advising parents and caregivers not to use these drugs in children younger than 2 [1]. This recommendation was based on information from ongoing studies reporting on safety concerns (such as cardiac arrhythmias, depressed levels of consciousness, encephalopathy and even death) of CCMs when used in young children [3, 12]. Following the US, warnings were as well issued in Canada, the United Kingdom and a partial warning in Italy on the use of nasal sympathomimetics in children <12 years. In the Netherlands, no national warning advising against the use of these drugs was spread. Although CCMs can be purchased as OTC in many countries, they are also regularly prescribed by primary care physicians, especially when these drugs are being reimbursed. We assessed the eff ect of these regulatory warnings on prescription rates of CCMs in children < 2 years in both Italy and the Netherlands (chapter 2.4). We found that, despite the international warnings and important publications in high impact journals and lay press advising against the use of CCMs in young children, overall prescription rates for CCMs increased in the Netherlands. In Italy, where a specifi c warning was issued against the use of nasal sympathomimetics or combinations, a signifi cant reduction in use of nasal combination products and nasal sympathomimetics was observed as well as a decrease in use of cough suppressants (opioids and non-opioids). The increase of the prescription of CCMs in the Netherlands is worrisome, especially since there is no evidence of effi cacy of these drugs [13]. The few studies that have been conducted in children found no signifi cant improvement in symptomatic relief when compared to placebo [13-17]. Caregiver education and physician reminding on the self-limiting nature of coughs and colds and on the lack of effi cacy of these CCMs is needed [3, 18]. We also would advice the Euro- pean Medicines Agency to explicitly stipulate the hazards of use of CCMs in young children.

Safety studies

Safety studies within Vigibase Although randomized controlled trials (RCTs) are considered to be the gold standard for performing safety studies, they are often hampered by small sample sizes and short follow-up. Spontaneous adverse event reporting databases, such as the Vigibase database of the WHO- UMC (Uppsala Monitoring Centre), contain the necessary data to elucidate potential safety risks of drugs among all age categories without the need to expose children to unnecessary clinical research. To our knowledge, Vigibase has not been previously used to describe pediatric adverse drug reactions (ADRs). In this thesis we present the fi rst studies on pediatric ADRs related to asthma medicines and cough and cold medicines using data from the Vigibase. We analysed all spontaneously reported ADRs for asthma and cough and cold medicines as captured in the Vigibase database of the WHO-UMC during the period 2000-2006. Associations between the use of these medicines and ADRs were assessed by calculating reporting odds

154

EEliflif FatmaFatma SSenen BW.inddBW.indd 115454 003-01-113-01-11 15:1715:17 Chapter 4: General Discussion

ratio (ROR) as a measure of disproportionality while adjusting for gender, continent and type of notifi er.

Asthma medicines related adverse drug reactions in children – data from the Vigibase database The risk-benefi t profi le of asthma medicines has been well studied in randomized controlled trials (RCTs), and is considered to be favorable [19]. However, the majority of these studies have been performed in children older than 5 years and studies in younger children are lacking. To get a broader understanding of the adverse drug reactions associated with asthma medicines use in children and to get a better insight in the potential risks of these medicines in children of all age categories, we analysed the spontaneously reported ADRs for asthma medicines in the Vigibase database. When the records in the VIGIBASE were limited to paediatric data, only 3.5% of the (non-vaccine related) records concerned ADRs related to the use of asthma medicines. This percentage does not correlate with the high prevalence of use of these medicines in children, e.g. 4-26% of children is being prescribed an asthma medicine (depending on country, age range, and study period) [1, 20]. This low reporting rate could be due to the fact that most asthma drugs are on the market for many years and treating physicians do not focus on potential (new) ADRs , because the medicines are thought to be safe. It could as well be that the low percentage of ADRs of asthma medicines, in relation to their high prevalence of use, is a true refl ection of the safety of these drugs. This is in contrast to the recent safety concerns on the use of e.g. LABA in children with asthma and the risk of serious asthma related events and/or death [21]. The majority of the associations that were observed are also described in literature and in the SPCs of the drugs of interest (e.g. adrenal insuffi ciency and use of ICS and Churg Strauss syndrome in relation to LTRA) [22, 23]. We observed very few ADRs with high ROR that were not yet described in literature. New ADRs with high ROR were heart valve disorder, intestinal gangrene and gastric ulcer perforation with use of short-acting beta-agonists, gastric ulcer perforation, hemiplegia and tooth disorder with use of long-acting beta-agonists and hypokinesia with use of xanthines. Although we did fi nd some new signals, we want to re-emphasize that ADRs from spontaneous reporting databases are purely hypothesis generating and further research is needed to assess these potentially new signals in an appropriate setting [24].

Cough and cold medicines related adverse drug reactions in children - data from the Vigibase database Because of the high use of CCMs and the concerns on the benefi t-risk profi le of these drugs in (young) children, we used the Vigibase database to investigate the type of adverse drug reactions and ROR that have been reported for CCMs in children aged 0-18 years.

155

EEliflif FatmaFatma SSenen BW.inddBW.indd 115555 003-01-113-01-11 15:1715:17 CCMs related ADRs were infrequently reported (1.2% of the non-vaccine related records), although studies in the USA have shown that in a given week, 10% of the children use a CCM [25]. Actual use is hard to estimate since many CCMs are sold over the counter, this may also have an impact on the reporting rate of spontaneous ADRs, since only few countries allow for direct consumer reporting. Similar to what was observed for the asthma medicines, it could as well be that CCMs are thought to be safe and potential ADRs are not recognized nor reported. Most of the reported ADRs for CCMs concerned mucolytics and opium alkaloids and derivatives. The majority of the observed associations between CCMs and ADRs in the Vigibase were already described in literature and SPCs (e.g. dosing errors and the typical side eff ects of opium alkaloids) [26, 27]. Two new and potentially life-threatening ADRs were observed namely chronic renal failure associated with the use of expectorants and cerebral infarction associated with the use of the fi xed combinations of opium alkaloids and expectorants. We want to re- emphasize that new CCMs-ADRs associations in spontaneous reporting databases are purely hypothesis generating and further research is needed to assess these potentially new signals in an analytical epidemiological approach. Overall, we found that for all age groups, various ADRs of diff erent levels of severity were reported. This shows that CCMs not only cause ADRs in the youngest children, but in children of all ages. We underline the advice not to use CCMs in young children as evidence of their effi cacy for the treatment of cough and cold is weak.

Glucocorticosteroids and risk of fractures According to the Global Initiative for Asthma (GINA), maintenance treatment with low dose ICS is the treatment of choice for children with persistent asthma [28]. Oral glucocorticoids (OCS) are recommended for the treatment of very severe persistent asthma exacerbations. In accordance with the guidelines, the use of ICS in asthmatic children is high [29]. Long term use of ICS in children has been associated with major safety concerns such as adrenal suppression and growth retardation [28]. Long term use of OCS can cause osteoporosis, diabetes mellitus, adrenal suppression, obesity, skin thinning leading to easy bruising, and muscle weakness [28]. The association between the use of ICS and the risk of fractures remains controversial and very few studies have been conducted in children. All of these studies have focused on one country at the time, and therefore were reduced in size and heterogeneity of exposure was low (United Kingdom and Australia). To our knowledge, there are no studies quantifying the association between corticosteroids use and risk of fracture in children, combining data from diff erent European countries, with diff erent user patterns. To get a better understanding of the risk of fractures associated with the use of corticosteroids in children, we conducted a population based nested case-control study using two primary care databases (IPCI and Pedianet) with quite diff erent prescription patterns of ICS and OCS in children and including children 0-18 years of age (0-14 for Pedianet). The case-control study was nested in a cohort of users of asthma medicines. We analysed the association between cumulative use of inhaled corticosteroids (ICS), oral corticosteroids (OCS), and use of a combination of

156

EEliflif FatmaFatma SSenen BW.inddBW.indd 115656 003-01-113-01-11 15:1715:17 Chapter 4: General Discussion

inhaled and oral corticosteroids (ICS+OCS) and the risk of fractures. No association was found between use of ICS or use of OCS and the risk of fractures in any of the exposure categories. None of the previously conducted studies in children found an increased risk of fractures following use of ICS [30-32]. In contrast to the studies from Van Staa et al., we did not fi nd an increased risk of fractures for children using OCS. Van Staa et al. found that the risk of fractures was increased in children receiving more than 3 prescriptions of OCS [33]. In our study, the majority of children used OCS for less than 30 days. Indeed only 1% of the children (both cases and controls) used OCS longer than one month. This duration is probably too short to increase the risk of fractures. Our fi ndings, as well as data from other studies, show that ICS, which are the mainstay of pediatric asthma treatment, can be used in children without concerns on a higher risk of fractures. We expect that long term use of OCS would increase the risk of fractures. As children are rarely chronically treated with OCS, this could not be studied in our database.

Glucocorticosteroids and risk of pancreatitis Pancreatitis is a relatively rare condition in children and is associated with considerable morbidity and mortality [34]. Pancreatitis mainly occurs in adult life and the mean age of a fi rst pancreatitis attack has been shown to be in the 6th decade of life [35]. In adults, the most common causes of acute pancreatitis are gallstones and alcohol abuse [35, 36]. In children, acute pancreatitis has diff erent aetiologies such as trauma, infection, systemic diseases, toxins, structural anomalies of the pancreatico-biliary tract and idiopathic [37-39]. Drug-induced pancreatitis is considered to be a rare cause of pancreatitis in both children and adults. Glucocorticosteroids are among the drugs that have been associated with pancreatitis, both in adults and children [38-41]. All of the existing data on this subject is derived from case reports and case series. To quantify the association between the use of glucocorticosteroids (both inhaled and oral) and pancreatitis in children, we conducted a population-based case-control study within the PHARMO record linkage system. In a source population of around half a million children we identifi ed only 54 children with acute pancreatitis, demonstrating the rarity of the condition. Many children with pancreatitis had been previously exposed to some type of drug. Current use of oral glucocorticosteroids increased the risk of pancreatitis with a factor 20 compared to no use (OR 21.40, p<0.001). Use of Inhaled glucocorticosteroids use was not associated with an increased risk. Our results are in line with case series in children, observing an association between oral glucocorticosteroids use and an increased risk of acute pancreatitis [39, 42, 43]. Similar to what has been described in literature, we confi rmed known risk factors of pancreatitis such as a history of infl ammatory bowel disease, gallstones, cystic fi brosis and trauma; and concomitant use of azathioprine, mesalazine, metronidazol, opiates, omeprazol and NSAIDs. [37, 44-48].

157

EEliflif FatmaFatma SSenen BW.inddBW.indd 115757 003-01-113-01-11 15:1715:17 General consideration regarding the use and safety of respiratory medications in children In this thesis we demonstrated that respiratory medicines are amongst the most frequently used drugs, particularly the inhaled corticosteroids and bronchodilating agents. These drugs merit their place for the treatment of asthma but in young children, these drugs are often used whereas the diagnosis of asthma could not be confi rmed. In some countries (e.g. Italy) use of oral steroids is quite high but often for a short duration of follow-up. We showed that each of the classes of respiratory drugs has ADRs reported to the Vigibase of the WHO-UMC. Most ADRs are already known in the literature or in the SPC, but some potential safety signals for these drugs were observed which require further formal testing. In particular cerebral infarction associated with use of opium alkaloids+expectorants, chronic renal failure associated with expectorants use, intestinal gangrene, heart valve disorders and gastric ulcer perforation with use of short-acting beta-agonists; and hemiplegia and tooth disorders with the use of long- acting beta-agonists need to be further explored, as these are serious events.

We quantifi ed the association for ICS and OCS with two particular (known) side eff ects (fracture and pancreatitis) in children. Use of ICS and OCS was not associated with an increased risk of fractures, whereas OCS, but not ICS, increased the risk of pancreatitis in children.

In the past; children were deprived of new drugs as, for ethical reasons, they were often excluded from clinical trials and thus effi cacy and safety data on use in children were lacking: The pediatric regulation aims to facilitate the development and accessibility of medicinal products for use in the pediatric population. For each dossier of a new drug; being submitted for approval, the law stipulates the need of a pediatric investigation plan (PIP): A PIP is a development plan aimed at ensuring that the necessary data are obtained through studies in children to support the authorisation of the medicine for children. The ultimate goal of the new regulation is to stimulate drug research in children to ensure the safety and effi cacy of drugs used in children. The implementation of this new regulation should be closely monitored as the drawback could be that children are exposed to unnecessary clinical trials when crucial data can be collected from other sources. Datasources that can be used for analysis are the electronic health care databases which contain information on the incidence/prevalence of the disease, the standard treatment of the disease and potential safety concerns. These are aspects that are part of the requested information from the PIP [49]. To convince skeptics of the importance and added value of observational databases, collaborative actions should be taken to further improve the quality of observational data. It is known that databases on spontaneous ADRs suff er from underreporting. Indeed, it has been estimated that less than 10% of all serious, and 2-4% of all non-serious ADRs are reported [50]. Reporting of spontaneous ADRs should be stimulated by adequate training and information programs, feed back systems and a facilitation of the reporting procedures. The research on rare side

158

EEliflif FatmaFatma SSenen BW.inddBW.indd 115858 003-01-113-01-11 15:1715:17 Chapter 4: General Discussion

eff ects and on long term side eff ects requires databases of considerable size with long term follow-up. This can be achieved by linking databases from various sources. Further research and development is needed to deal with issues such as database ownership, and transcription to uniform disease or drug coding. In addition, databases should try to capture data on important confounding factors. With regard to the safety and eff ectiveness of respiratory drugs, factors that need to be captured are asthma phenotypes and asthma severity as they modify treatment response. In our research on the safety of respiratory drugs, we fi rst checked potential safety signals using spontaneous ADR data from the Vigibase database. In addition, we quantifi ed the association between the use of glucocorticosteroids and the risk of fractures and pancreatitis. We are aware that there are currently other safety concerns in relation to respiratory drugs that were not further elucidated. These concerns are the use of LABA in children and the risk of serious asthma related events and/or cardiovascular events. These safety concerns were not further explored for reasons of low LABA exposure and the very low incidence of cardiovascular events in children.

Methodological considerations In this thesis we have used several readily available healthcare databases to perform large scale pediatric observational studies. The studies performed in this thesis are part of the TEDDY NoE, which aims to promote the availability of safe and eff ective medicines for children in Europe by integrating existing expertise and good practices To our knowledge, we are the fi rst to combine data from health care databases by using a predefi ned, common protocol. Combining the databases in this manner allowed us to perform large-scale pharmacoepidemiological research in pediatrics to get better knowledge about the safety and potentially benefi ts of drugs in children, without exposing children to unnecessary RCTs.

Primary care physician databases For several studies in this thesis we used the the Pedianet database (pediatric electronic medical records from 150 pediatricians since 2000) in Italy [51], the IPCI database (comprising adult and pediatric electronic medical records from more than 400 doctors since 1996) in The Netherlands [52, 53], and the IMS DA database (electronic medical records on adults and children from 670 doctors) in the UK [54] . All of these databases contain the complete automated medical records of primary care physicians and have been used and proven valid for pharmacoepidemiological research [55]. The age and gender distribution in the various databases is representative for the country they originate from. In these countries, primary care physicians are the guardians of children’s health, which means that all clinical information concerning the patients (including summaries of specialist and hospital care) is kept in their medical records. Since all children need to be registered with a GP

159

EEliflif FatmaFatma SSenen BW.inddBW.indd 115959 003-01-113-01-11 15:1715:17 in the Netherlands and United Kingdom, and with a family pediatrician in Italy, the databases are population-based [55].

Limitations of these databases in pharmacoepidemiological studies

Exposure, information and selection bias One limitation of these databases is that they only capture outpatient, primary care prescriptions and no specialist prescriptions nor over the counter drugs (OTC). In addition, as we only had access to prescription data, information on actual refi ll of the prescriptions was missing meaning that we overestimated the actual exposure. In the studies on medicines use overall and use of CCMs, this may have resulted in underestimation of prevalence of use of medicines which are commonly used OTC (e.g. paracetamol, xylometazoline). In our study on asthma medicines use, this will minimally infl uence our results since asthma is a condition often dealt with in primary care, and asthma medicines originally prescribed by a specialist are often continued or prescribed by general practitioners [53].

In our case-control study on the use of glucocorticosteroids and the risk of fractures, we do believe that misclassifi cation of exposure minimally infl uenced our exposure rates as asthma is a condition often dealt with in primary care, and asthma medicines originally prescribed by a specialist are often continued or prescribed by general practitioners [53]. If present; this misclassifi cation is probably non-diff erential between cases and controls, therefore the actual risk may have been underestimated.

Selection bias was unlikely, as all data was obtained from prospectively collected medical records that are maintained for patients care purposes.

Information bias via misclassifi cation of the outcome is also unlikely, because all fracture cases were retrieved from medical records and reviewed by two medical experts who were blinded to the exposure.

Confounding A confounder is a factor that is both associated with the disease of interest and associated with the exposure but is not part of the common pathway (Figure 1). Due to the presence of the confounder, it is not possible to accurately assess the relationship between the exposure and outcome of interest. Confounding may be accounted for by matching participants by likely confounders (such as age and sex), or controlling the eff ect of the confounder by stratifying the analysis.

160

EEliflif FatmaFatma SSenen BW.inddBW.indd 116060 003-01-113-01-11 15:1715:17 Chapter 4: General Discussion

We used IPCI and Pedianet to assess the association between steroid use and risk of fractures. Although we tried to adjust for any potential confounders, due to the nature of the databases, we did not have information on all possible confounders such as the nutritional status of the children, body mass index, bone mineral density, physical activity and smoking status. Residual confounding usually results in an overestimation of the observed risk. However, as we did not observe an increased risk we do not believe that the lack of information on these confounders infl uences our results.

Strengths A major strength of these databases is that they capture a large number of children in databases across Europe, resulting in a high generalisability of our fi ndings to the general population. In addition, these databases have detailed information on comorbidity and detailed information on hospital referral and discharge letters. This results in a good quality of case ascertainment. The fact that data on millions of children is available in health care databases should be recognized by researchers and eff orts should be made to extract data on pediatric drug use/safety from these databases, instead of performing unnecessary clinical trials.

PHARMO RLS We used PHARMO RLS to assess the association between the use of glucocorticosteroids and risk of pancreatitis.

Exposure and information bias In contrast to the IPCI and IMS data, the PHARMO-RLS captures refi ll data. Still, exposure bias might be present as the database does not capture actual intake of the drug. If present; we do believe that this bias will be non-diff erential as well, leading to an underestimation of the actual risk.

Information bias by misclassifi cation of the outcome may be an issue using the PHARMO database. All outcomes are coded by ICD-9 codes, and manual validation of potential cases is not possible as it would require the de-anonymisation of hospital data and permission to review the fi les.

In addition, we may suff er from protopathic bias since steroids are also given to treat pancreatitis. This could not be the explanation of our fi ndings since the association was only observed in children using oral steroids for 7 days or more.

Confounding

161

EEliflif FatmaFatma SSenen BW.inddBW.indd 116161 003-01-113-01-11 15:1715:17 A major limitation of this study is that, as with each observational retrospective study, the fi ndings were prone to unmeasured confounding as information on body mass index and alcohol intake were missing.

Strength The Pharmo RLS has as strength that it is a very large database containing information on more than 3 million inhabitants from the Netherlands meaning that our results can be extrapolated to the entire pediatric population of the Netherlands. In addition, as PHARMO contains pharmacy dispensing data, both prescriptions from general practitioners and specialists were captured, ensuring a more complete overview of the drugs used. Although the database is large, pancreatitis is very rare and we this lacked power. Future studies should use approaches as done in TEDDY: pooling across databases.

Vigibase Vigibase was used to provide an overview of pediatric ADRs related to asthma and cough and cold medicines. Data from spontaneous reporting databases suff er from bias such as under and overreporting. [56-58]. Furthermore, is has to be underlined that a signal with a high ROR does not prove that the drug and ADR are causally related. The observed signals are purely hypothesis generating and further research is needed to assess these signals in an appropriate setting [24]. Also, some of the found signals are due to confounding by indication, such as cystic fi brosis associated to mucolytics. Due to the nature of the database, information on indication is missing, and therefore it is not possible to identify the underlying condition. The strength of this database is the large number of pediatric ADRs available for analysis.

Recommendations for future research The studies in this thesis may help to give direction to further research in the fi eld of pediatric medicines research. First, primary care physicians should be aware of the harms of CCMs in children. The fact that prescriptions rates of CCMs in the Netherlands even increased after the international warnings is alarming. We believe that in Europe, this topic did not receive the attention it deserves. We highly recommend that a concerted European action is undertaken to advice against the use and prescription of cough and cold medicines to children.

Second, we have shown the potential of linking multi-country databases to study country specifi c pediatric drug use and safety in a systematic manner without being hampered by methodological diff erences. Although RCTs are the gold standard for effi cacy assessment, safety studies are more eff ectively conducted through large scale postmarketing observational studies [59]. Clinical trials are often small and too short for safety assessment

162

EEliflif FatmaFatma SSenen BW.inddBW.indd 116262 003-01-113-01-11 15:1715:17 Chapter 4: General Discussion

Pharmacoepidemiological studies should be promoted to improve the safety assessment of pediatric drugs since they assess the safety of drugs under real life circumstances. As safety information is lacking on long-term use of controller medication in asthmatic children, healthcare databases throughout Europe should be combined to perform large scale safety studies. Also, the ADRs we found within Vigibase, which were not yet described in literature, can be assessed in this manner. Combining several databases will make it possible to also detect rare ADRs.

This thesis shows that exposure and outcome information is available on a large pediatric population within Europe. Researchers should try to work together on linking the available databases within Europe, instead of performing clinical trials on a small scale.

163

EEliflif FatmaFatma SSenen BW.inddBW.indd 116363 003-01-113-01-11 15:1715:17 References

1 Sturkenboom MC, Verhamme KM, Nicolosi A, Murray ML, Neubert A, Caudri D, et al. Drug use in children: cohort study in three European countries. BMJ (Clinical research ed. 2008;337:a2245. 2 List of paediatric needs (as established by the Paediatric Working Party). [cited; Available from: http:// www.emea.europa.eu/htms/human/paediatrics/inventory.htm 3 Sharfstein JM, North M, Serwint JR. Over the counter but no longer under the radar--pediatric cough and cold medications. N Engl J Med. 2007 Dec 6;357(23):2321-4. 4 FDA. FDA Recommends that Over-the-Counter (OTC) Cough and Cold Products not be used for Infants and Children under 2 Years of Age. 2008 [cited 03-10-2009]; Available from: http://www.fda. gov/Drugs/DrugSafety/PublicHealthAdvisories/ucm051137.html 5 Neubert A, Bonifazi A, Catapano M, Baiardi P, Guiaquinto C, Knibbe C, et al. Defi ning Off -label and Unlicensed Use of Medicines for Children: Results of a Delphi survey. Pharmacol Res. 2008;submitted. 6 Clavenna A, Bonati M. Drug prescriptions to outpatient children: a review of the literature. European journal of clinical pharmacology. 2009 Aug;65(8):749-55. 7 Heinemann L, Lewis M, Spitzer W, Thorogood M, Guggenmoos-Holzmann I, Bruppacher R. Throm- boembolic stroke in young women. A European case-control study on oral contraceptives. Trans- national Research Group on Oral Contraceptives and the Health of Young Women. . Contraception. 1998;57:29-37. 8 Heinemann L, Lewis M, Thorogood M, Spitzer W, Guggenmoos-Holzmann I, Bruppacher R. Case- control study of oral contraceptives and risk of thromboembolic stroke: results from International Study on Oral Contraceptives and Health of Young Women. BMJ. 1997;315:1502-4. 9 Lewis M. Myocardial infarction and stroke in young women: what is the impact of oral contraceptives? Am J Obstet Gynecol. 1998;179:S68-77. 10 Samuelsson E, Hagg S, . Incidence of venous thromboembolism in young Swedish women and possibly preventable cases among combined oral contraceptive users. . Acta Obstet Gynecol Scand. 2004;83::674-81. 11 Samuelsson E, Hedenmalm K, Persson I. Mortality from venous thromboembolism in young Swedish women and its relation to pregnancy and use of oral contraceptives--an approach to specifying rates. . Eur J Epidemiol 2005;20:509-16. 12 Infant deaths associated with cough and cold medications--two states, 2005. Mmwr. 2007 Jan 12;56(1):1-4. 13 Smith SM, Schroeder K, Fahey T. Over-the-counter medications for acute cough in children and adults in ambulatory settings. Cochrane database of systematic reviews (Online). 2008(1):CD001831. 14 Smith MB, Feldman W. Over-the-counter cold medications. A critical review of clinical trials between 1950 and 1991. Jama. 1993 May 5;269(17):2258-63. 15 Taylor JA, Novack AH, Almquist JR, Rogers JE. Effi cacy of cough suppressants in children. The Journal of pediatrics. 1993 May;122(5 Pt 1):799-802. 16 Paul IM, Yoder KE, Crowell KR, Shaff er ML, McMillan HS, Carlson LC, et al. Eff ect of dextromethorphan, diphenhydramine, and placebo on nocturnal cough and sleep quality for coughing children and their parents. Pediatrics. 2004 Jul;114(1):e85-90. 17 Paul IM, Beiler J, McMonagle A, Shaff er ML, Duda L, Berlin CM, Jr. Eff ect of honey, dextromethorphan, and no treatment on nocturnal cough and sleep quality for coughing children and their parents. Archives of pediatrics & adolescent medicine. 2007 Dec;161(12):1140-6. 18 Use of codeine- and dextromethorphan-containing cough remedies in children. American Academy of Pediatrics. Committee on Drugs. Pediatrics. 1997 Jun;99(6):918-20.

164

EEliflif FatmaFatma SSenen BW.inddBW.indd 116464 003-01-113-01-11 15:1715:17 Chapter 4: General Discussion

19 Graham LM. Balancing safety and effi cacy in the treatment of pediatric asthma. The Journal of allergy and clinical immunology. 2002 Jun;109(6 Suppl):S560-6. 20 Bianchi M, Clavenna A, Bonati M. Inter-country variations in anti-asthmatic drug prescriptions for children. Systematic review of studies published during the 2000-2009 period. European journal of clinical pharmacology. Sep;66(9):929-36. 21 US Food and Drug Administration. FDA Drug Safety Communication: New Safety Requirements for Long-Acting Inhaled Asthma Medications Called Long-Acting Beta-Agonists (LABAs). 18-02-2010 [cited; Available from: http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationfor- PatientsandProviders/ucm200776.htm 22 Electronic Medicines Compendium. Summary of Product Characteristics for Easyhaler Budesonide 400mcg, last updated on the eMC: 02/07/2009. 2010 [cited; Available from: http://www.medicines. org.uk/EMC/medicine/18774/SPC/Easyhaler+Budesonide+400mcg/ 23 Electronic Medicines Compendium. Summary of Product Characteristics for SINGULAIR PAEDIATRIC 4 mg GRANULES, last updated on the eMC: 02/07/2010. 2010 [cited; Available from: http://www. medicines.org.uk/EMC/medicine/14071/SPC/SINGULAIR+PAEDIATRIC+4+mg+GRANULES/ 24 Almenoff J, Tonning JM, Gould AL, Szarfman A, Hauben M, Ouellet-Hellstrom R, et al. Perspectives on the use of data mining in pharmaco-vigilance. Drug Saf. 2005;28(11):981-1007. 25 Vernacchio L, Kelly JP, Kaufman DW, Mitchell AA. Cough and cold medication use by US children, 1999-2006: results from the slone survey. Pediatrics. 2008 Aug;122(2):e323-9. 26 Dart RC, Paul IM, Bond GR, Winston DC, Manoguerra AS, Palmer RB, et al. Pediatric fatalities associated with over the counter (nonprescription) cough and cold medications. Annals of emergency medicine. 2009 Apr;53(4):411-7. 27 Olson KR. Poisoning & Drug Overdose, Fourth Edition. 2004 [cited May 26, 2010]; Available from: http://www.hrd.gov.co/INTOXICACION/Poisoning%20and%20Drug%20Overdose.pdf 28 The Global Strategy for Asthma Management and Prevention, Global Initiative for Asthma (GINA) 2007. [cited; Available from: http://www.ginasthma.org 29 Sen EF, Verhamme KM, Neubert A, Hsia Y, Murray M, Felisi M, et al. Assessment of Pediatric asthma drug use in three European countries; a TEDDY study. European journal of pediatrics. Sep 2. 30 Schlienger RG, Jick SS, Meier CR. Inhaled corticosteroids and the risk of fractures in children and adolescents. Pediatrics. 2004 Aug;114(2):469-73. 31 van Staa TP, Bishop N, Leufkens HG, Cooper C. Are inhaled corticosteroids associated with an increased risk of fracture in children? Osteoporos Int. 2004 Oct;15(10):785-91. 32 Ma DQ, Jones G. Clinical risk factors but not bone density are associated with prevalent fractures in prepubertal children. Journal of paediatrics and child health. 2002 Oct;38(5):497-500. 33 van Staa TP, Cooper C, Leufkens HG, Bishop N. Children and the risk of fractures caused by oral corticosteroids. J Bone Miner Res. 2003 May;18(5):913-8. 34 Corfi eld AP, Cooper MJ, Williamson RC. Acute pancreatitis: a lethal disease of increasing incidence. Gut. 1985 Jul;26(7):724-9. 35 Yadav D, Lowenfels AB. Trends in the epidemiology of the fi rst attack of acute pancreatitis: a systematic review. Pancreas. 2006 Nov;33(4):323-30. 36 Whitcomb DC. Clinical practice. Acute pancreatitis. The New England journal of medicine. 2006 May 18;354(20):2142-50. 37 Yeung CY, Lee HC, Huang FY, Ho MY, Kao HA, Liang DC, et al. Pancreatitis in children--experience with 43 cases. European journal of pediatrics. 1996 Jun;155(6):458-63. 38 Alvarez Calatayud G, Bermejo F, Morales JL, Claver E, Huber LB, Abunaji J, et al. Acute pancreatitis in childhood. Rev Esp Enferm Dig. 2003 Jan;95(1):40-4, 5-8.

165

EEliflif FatmaFatma SSenen BW.inddBW.indd 116565 003-01-113-01-11 15:1715:17 39 Vane DW, Grosfeld JL, West KW, Rescorla FJ. Pancreatic disorders in infancy and childhood: experience with 92 cases. Journal of pediatric surgery. 1989 Aug;24(8):771-6. 40 Riemenschneider TA, Wilson JF, Vernier RL. Glucocorticoid-induced pancreatitis in children. Pediatrics. 1968 Feb;41(2):428-37. 41 Weaver G. Steroid-induced pancreatitis. Gastroenterology. 1982 Mar;82(3):601. 42 Barr HS, Wolff OH. Pancreatic necrosis in cortisone-treated children. Lancet. 1957 Apr 20;272(6973): 812-5. 43 Marczynska-Robowska M. Pancreatic necrosis in a case of Still’s disease. Lancet. 1957 Apr 20;272(6973):815-6. 44 Hillemeier C, Gryboski JD. Acute pancreatitis in infants and children. The Yale journal of biology and medicine. 1984 Mar-Apr;57(2):149-59. 45 Weizman Z, Durie PR. Acute pancreatitis in childhood. The Journal of pediatrics. 1988 Jul;113(1 Pt 1):24-9. 46 Balani AR, Grendell JH. Drug-induced pancreatitis : incidence, management and prevention. Drug Saf. 2008;31(10):823-37. 47 Trivedi CD, Pitchumoni CS. Drug-induced pancreatitis: an update. Journal of clinical gastroenterology. 2005 Sep;39(8):709-16. 48 Nitsche CJ, Jamieson N, Lerch MM, Mayerle JV. Drug induced pancreatitis. Best practice & research. Apr;24(2):143-55. 49 European Medicines Agency, Paediatric investigation plans (PIPs), waivers and modifi cations. [cited; Available from: http://www.ema.europa.eu/ema/index.jsp?curl=pages/regulation/document_list- ing/document_listing_000293.jsp&murl=menus/regulations/regulations.jsp&mid=WC0b01ac05800 25b91&jsenabled=true 50 Pirmohamed M, Breckenridge AM, Kitteringham NR, Park BK. Adverse drug reactions. BMJ (Clinical research ed. 1998 Apr 25;316(7140):1295-8. 51 Sturkenboom M, Nicolosi A, Cantarutti L, Mannino S, Picelli G, Scamarcia A, et al. Incidence of muco- cutaneous reactions in children treated with nilfumic acid, other nonsteroidal antiinfl ammatory drugs, or nonopioid analgesics. Pediatrics. 2005;116:e26-33. 52 Vlug A, van der Lei J, Mosseveld B, van Wijk M, van der Linden P, Sturkenboom M, et al. Postmarketing surveillance based on electronic patient records: the IPCI project. Methods Inf Med. 1999;38:339-44. 53 van der Lei J, Duisterhout J, Westerhof H, van der Does E, Cromme P, Boon W, et al. The introduction of computer-based patient records in the Netherlands. Ann Intern Med. 1993;119(10):1036-41. 54 Wong I, Murray M. The potential of UK clinical databases in enhancing paediatric medication research. Br J Clin Pharmacol. 2005;59(6):750-5. 55 Sturkenboom M. Other European databases for pharmacoepidemiology. . In: Mann RD AE, ed. Phar- macovigilance`. second ed. London: Wiley 2007. 56 Hartnell NR, Wilson JP. Replication of the Weber eff ect using postmarketing adverse event reports voluntarily submitted to the United States Food and Drug Administration. Pharmacotherapy. 2004 Jun;24(6):743-9. 57 Motola D, Vargiu A, Leone R, Conforti A, Moretti U, Vaccheri A, et al. Infl uence of regulatory measures on the rate of spontaneous adverse drug reaction reporting in Italy. Drug Saf. 2008;31(7):609-16. 58 Pariente A, Gregoire F, Fourrier-Reglat A, Haramburu F, Moore N. Impact of safety alerts on measures of disproportionality in spontaneous reporting databases: the notoriety bias. Drug Saf. 2007;30(10):891-8. 59 Sorensen HT, Lash TL, Rothman KJ. Beyond randomized controlled trials: a critical comparison of trials with nonrandomized studies. Hepatology (Baltimore, Md. 2006 Nov;44(5):1075-82.

166

EEliflif FatmaFatma SSenen BW.inddBW.indd 116666 003-01-113-01-11 15:1715:17 Summary

EEliflif FatmaFatma SSenen BW.inddBW.indd 116767 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 116868 003-01-113-01-11 15:1715:17 Summary

The Task-force in Europe for Drug Development for the Young (TEDDY) Network of Excellence was established in 2005 through funding of the European Community’s sixth Framework Programme to contribute to the promotion of safe and effi cacious medicines for children in the context of the impending European Paediatric Regulation that came into force in January 2007. The overall aim of the TEDDY NoE is to promote the availability of safe and eff ective medicines for children in Europe by integrating existing expertise and good practices, as well as stimulating pediatric drug development. The work presented in this thesis is a spin-off of the TEDDY activities. In this thesis, we aimed to study the use and safety of two mostly used medicine groups in children, namely asthma medicines and cough and cold medicines. We did this through large scale observational studies, by combining readily available healthcare databases within Europe.

In chapter 2 we present several multi-national studies on the utilization of medicines in children. In chapter 2.1 and 2.2 we present the utilization of medicines in children living in Italy, the Netherlands and the United Kingdom. Children up to 14 (Italy) or 18 years of age (United King- dom and The Netherlands) from three primary care databases were included in the studies. These were the fi rst studies describing pediatric medicines use in diff erent European countries by combining databases using the same methodology Prevalence of use per year was calculated by drug class (anatomic and therapeutic levels). A total of 675,868 children were included in the studies. The mostly used medicines were anti-infectives, dermatological and respiratory medicines, whereas cardiovascular and anti- neoplastic medicines had the lowest use.

In chapter 2.3 we describe the utilization of asthma medicines in children from Italy, the Netherlands and the United Kingdom by combining health care databases. For all children, prescription rates of asthma drugs were studied by country, age, asthma diagnosis and off - label status. The cohort consisted of 671,831 children of whom 49,442 had been diagnosed with asthma at any time during follow-up. ß2-mimetics and inhaled steroids were the most frequently prescribed asthma medicine classes in the studied countries. Xanthines, anticholinergics, leukotriene receptor antagonists and anti-allergics were the least frequently prescribed medicines classes. Off -label use was low, and most pronounced for ß2-mimetics in children < 18 months (Italy) and combined ß2-mimetics+anticholinergics in children <6 years (Netherlands). With this study we showed that linking multi-country databases enables us study country specifi c pediatric drug use in a systematic manner without being hampered by methodological diff erences.

In chapter 2.4 we evaluate the eff ect of international safety warnings on the prescription rates of cough and cold medicines in children in the Netherlands and Italy. Cough and cold medicines are frequently used in children to treat upper respiratory tract infections without

169

EEliflif FatmaFatma SSenen BW.inddBW.indd 116969 003-01-113-01-11 15:1715:17 solid proof of benefi ts and risk of serious adverse events. In 2007 international warnings have been issued advising against use of these drugs in young children. We analysed outpatient electronic medical records of children <2 years in Italy and the Netherlands and calculated pre- and post-warning prescription rates. The cohort consisted of 99,176 children < 2 years of age. After international warnings, overall prescription rates for cough and cold medicines decreased slightly in Italy and increased in the Netherlands. Despite the international safety warnings and negative benefi t-risk profi les, prescription rates of cough and cold medicines remained substantial and were hardly aff ected by the warnings, especially in the Netherlands.

In chapter 3 we assess the safety of asthma and cough and cold medicines by analysing the adverse drug reactions (ADRs) from the WHO Vigibase database and by performing case- control studies. Chapters 3.1 and 3.2 give an overview of adverse events related to asthma medicines and cough and cold medicines, reported to Vigibase. Of the non-vaccine related records (294,773), 3.5% comprised of asthma medicines related ADRs and 1.2% of cough and cold medicines. For asthma medicines, the majority of the records were for children aged 2-≤11 years and most records concerned short-acting β2-agonists (SABAs) and inhaled corticosteroids. For cough and cold medicines, most ADRs were reported for children aged 2-18 years and mostly concerned mucolytics and opium alkaloids. In both studies, the majority of the found drug- ADR combinations were already described in literature. New, potential signals that we found Summary were were intestinal gangrene and perforated gastric ulcer associated with use of SABAs, tooth discoloration with the use of anticholinergics, chronic renal failure associated with use of expectorants and cerebral infarction associated with the use of opium alkaloids+expectorants.

In chapter 3.3 we describe a study on the association between use of corticosteroids (inhaled [ICS or oral [OCS]) and risk of fractures in children. Long term use of OCS has been associated with an increased risk of fractures both in children and adults; whereas the association between the use of ICS and the risk of fractures remains controversial. To assess the risk of fractures associated with the use of corticosteroids in children, we conducted a case-control study combining data from the Netherlands and Italy. We confi rmed known risk factors for fractures, such as trauma and current use of drugs for functional gastrointestinal disorders. For exposure to corticosteroids (ICS or OCS), we found no increased risk of fractures.

In chapter 3.4 a study on the association between use of corticosteroids (inhaled [ICS or oral [OCS]) and risk of pancreatitis in children is described. Although drug-induced pancreatitis is rare, a variety of drugs have been associated to pancreatitis (in children and adults), including steroids. So far, no studies have been performed to assess the association between exposure to steroids and pancreatitis in children. We conducted a case-control study within a Dutch hospital

170

EEliflif FatmaFatma SSenen BW.inddBW.indd 117070 003-01-113-01-11 15:1715:17 Summary

discharge registry. Current use of OCS and cumulative use of >7 days OCS was associated with a higher risk of pancreatitis. Use of ICS was not associated with pancreatitis.

Chapter 4 includes a general discussion in which the results and conclusions of studies in this thesis are summarized and interpreted. Furthermore, methodological considerations are discussed and suggestions are given for further research.

171

EEliflif FatmaFatma SSenen BW.inddBW.indd 117171 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 117272 003-01-113-01-11 15:1715:17 Sam envatting

EEliflif FatmaFatma SSenen BW.inddBW.indd 117373 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 117474 003-01-113-01-11 15:1715:17 Samenvatting

Gesubsidieerd door de Europese Commissie’s zesde Framework Programma is in 2005 de Task-force in Europe for Drug Development for the Young (TEDDY) Netwerk van Excellentie opgericht om een bijdrage te leveren aan de bevordering van veilige en werkzame medicijnen voor kinderen, in de context van de Europese Pediatrische Wetgeving, welke in Januari 2007 van kracht werd. Het algemene doel van TEDDY is het bevorderen van de beschikbaarheid van veilige en werkzame medicijnen voor kinderen in Europa, door zowel bestaande expertise samen te brengen als ook medicijn ontwikkeling specifi ek gericht op kinderen te stimuleren. De studies die gepresenteerd worden in dit proefschrift zijn het resultaat van TEDDY werkzaamheden. Het specifi eke doel van dit proefschrift was het onderzoeken van het gebruik en de veiligheid van twee veel gebruikte medicijn groepen in kinderen, namelijk astma medicijnen en Verkoudheidsmedicijnen. Dit hebben wij gedaan door medische databasen binnen Europa te combineren en hiermee observationele studies op grote schaal uit te voeren. In hoofdstuk 2 presenteren we meerdere multinationale studies naar het gebruik van medicijnen in kinderen. In hoofdstuk 2.1 en 2.2 beschrijven we het gebruik van medicijnen door kinderen in Italië, Nederland en het Verenigd Koninkrijk. Voor deze studies hebben we gegevens van kinderen uit drie huisarten-databasen gecombineerd. Deze studies waren de eerste in hun soort, waarin het medicijn gebruik door kinderen uit verschillende Europese landen wordt beschreven door medische databasen uit deze landen op uniforme wijze te combineren. Het gebruik van medicijnen per jaar werd berekend per medicijn-groep (op anatomisch en the- rapeutisch niveau). In totaal werden 675,868 kinderen geincludeerd in deze studies. De meest gebruikte medicijnen waren anti-infectiva, dermatologische preparaten en respiratoire medicijnen. Cardiovasculaire medcijnen en chemotherapeutica werden het minst vaak gebruikt. In hoofdstuk 2.3 beschrijven we het gebruik van astma medicijnen door kinderen in Italië, Nederland en het Verenigd Koninkrijk aan de hand van het aantal voorgeschreven recepten. Recepten voor astma medicijnen werden bestudeerd per land, leeftijd, astma diagnose en off -label status. In totaal werden 671,831 geincludeerd in deze studie, van welke 49,442 waren gediagnosticeerd met astma. β-sympathicomimetica en inhalatiecorticosteroïden waren de meest voorgeschreven astma medicijnen. Xanthines, anticholinergica, leukotriene-receptor antagonisten en anti-allergica werden het minst vaak voorgeschreven. Off -label gebruik van astma medicatie was laag, de middelen die het vaakst off -label werden gebruikt waren: β-sympathicomimetica door kinderen <18 maanden oud (in Italië) en β-sympathicomimetica+ anticholinergica door kinderen <6 jaar (in Nederland). Met deze studie tonen wij aan dat het combineren van multionationale medische databasen wetenschappers in staat stelt om medicijn gebruik door kinderen te bestuderen, zonder gehin- derd te worden door methodologische problemen.

175

EEliflif FatmaFatma SSenen BW.inddBW.indd 117575 003-01-113-01-11 15:1715:17 In hoofdstuk 2.4 wordt het eff ect van internationale veiligheidswaarschuwingen op het aantal recepten voor verkoudheidsmedicijnen aan kinderen in Nederland en Italië beschreven. Verkoudheidsmedicijnen worden vaak gegeven bij bovenste luchtweginfecties, hoewel hun werkzaamheid in kinderen nooit is bewezen en er een verhoogd risico op ernstige bijwerkingen is. Als gevolg hiervan, zijn er in 2007 internationale waarschuwingen gegeven, waarin werd geadviseerd deze medicijnen niet te geven aan jonge kinderen. Wij hebben de medische gegevens van kinderen <2 jaar in Italië en Nederland bestudeerd en het aantal voorschriften voor verkoudheidsmedicijnen voor en na de waarschuwingen bekeken. 99,176 kinderen jonger dan 2 jaar werden geïncludeerd in deze studie. Na de internationale waarschuwingen, daalde het aantal voorschriften voor verkoudheidsmiddelen in Italië licht en in Nederland steeg het aantal juist. Ondanks de internationale veiligheidswaarschuwingen én het risico op ernstige bijwerkingen, werden verkoudheidsmiddelen nog aanzienlijk voorgeschreven, vooral in Nederland. In hoofdstuk 3 bestuderen we de veiligheid van astma medicijnen en verkoudheidsmedicijnen door bijwerkingen in de Vigibase database te bestuderen en door case-controle studies uit te voeren. In hoofdstukken 3.2 en 3.2 wordt een overzicht gegeven van de bijwerkingen van astma medicijnen en verkoudheidsmedicijnen, die gemeld zijn aan de Vigibase database. Van alle bijwerkingen die niet aan vaccins waren gerelateerd (294,773), betrof 3.5% astma medicijnen en 1.2% verkoudheidsmedicijnen. De meeste bijwerkingen voor astma medicijnen werden Samenvatting gemeld voor kinderen tussen 2-≤11 jaar en betroff en voornamelijk kortwerkende β –sympathicomimetica en inhalatiecorticosteroïden. Voor verkoudheidsmedicijnen werden de meeste bijwerkingen gemeld voor kinderen tussen 2-18 jaar en betroff en voornamelijk mucolytica en opium alkaloïden. De meeste van de beschreven bijwerkingen-geneesmiddel combinaties in beide studies zijn gekend in de literatuur. Mogelijk nieuwe signalen die wij hebben gevonden zijn de volgende: gangreen van de darmen en perforatie van een maagzweer bij gebruik van kortwerkende β –sympathicomimetica, verkleuring van de tanden bij gebruik van anticholinergica, chronisch nierfalen bij gebruik van expectorantia en hersen-infarct bij gebruik van opium alkaloïden+expectorantia. In hoofdstuk 3.3 beschrijven we een studie naar de associatie tussen het gebruik van cortico- steroïden (zowel inhalatie als oraal) en fracturen. Langdurig gebruik van orale corticosteroïden is in de literatuur geassocieerd met een verhoogd risico op fracturen in zowel kinderen als volwassenen, het risico bij inhalatie corticosteroïden is tot op heden twijfelachtig. Om de associatie tussen corticosteroïden en fracturen te bestuderen in kinderen, hebben wij een case-controle studie uitgevoerd waarin data uit Nederland en Italië werden gecombineerd. Gekende risicofactoren voor fracturen, zoals trauma en het gebruik van medicatie voor gastro- intestinale stoornissen, werden bevestigd in onze studie. Het gebruik van corticosteroïden (zowel inhalatie als oraal) gaf geen verhoogd risico op fracturen in kinderen.

176

EEliflif FatmaFatma SSenen BW.inddBW.indd 117676 003-01-113-01-11 15:1715:17 Samenvatting

Hoofdstuk 3.4 beschrijft een studie naar het gebruik van corticosteroïden (zowel inhalatie als oraal) en het risico op pancreatitis bij kinderen. Hoewel pancreatitis veroorzaakt door medicijnen zeldzaam is, zijn een aantal medicijnen in verband gebracht met pancreatitis (zowel in kinderen als in volwassenen), waaronder steroïden. Tot op heden, zijn er geen studies uitgevoerd om het risico op pancreatitis bij gebruik van steroïden te beschrijven in kinderen. Om dit gat aan informatie te vullen, hebben wij een case-controle studie uitgevoerd, gebruik makend van de Nederlandse Pharmo database. Huidig gebruik van orale corticosteroïden en cumulatief gebruik van > 7 dagen waren geassocieerd met een verhoogd risico op pancreatitis. Het gebruik van inhalatie corticosteroïden gaf geen verhoogd risico. Hoofdstuk 4 bevat een algemene discussie, waarin de resultaten en conclusies van studies in dit proefschrift worden samengevat en geïnterpreteerd. Daarnaast worden methodologische aspecten besproken en worden suggesties gedaan voor verder onderzoek.

177

EEliflif FatmaFatma SSenen BW.inddBW.indd 117777 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 117878 003-01-113-01-11 15:1715:17 PhD po rtfolio

EEliflif FatmaFatma SSenen BW.inddBW.indd 117979 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 118080 003-01-113-01-11 15:1715:17 PhD portfolio

PhD Portfolio

Research skills

2007-2008 Master of Science in Clinical Epidemiology, Netherlands Institute for Health Sciences, Erasmus University, Rotterdam

International conference presentations

2008 11th Biannual Congress of the European Society for Developmental, Perinatal and Paediatric Pharmacology, Rotterdam, the Netherlands; poster presentation ‘Respiratory drug use in children in the Netherlands, Italy and United Kingdom’ .

2008 11th Biannual Congress of the European Society for Developmental, Perinatal and Paediatric Pharmacology, Rotterdam, the Netherlands; poster presentation ‘Use of cough and cold medications in children in the Netherlands and Italy’.

2008 24th International Conference on Pharmacoepidemiology, Copenhagen, Den- mark; poster presentation ‘Respiratory drug use in children in the Netherlands, Italy and United Kingdom’ .

2008 24th International Conference on Pharmacoepidemiology, Copenhagen, Den- mark; poster presentation ‘Use of cough and cold medications in children in the Netherlands and Italy’ .

2009 PRIOMED matchmaking event, Tallinn, Estonia; poster presentation ‘Cough and Cold medicine prescriptions in children’.

181

EEliflif FatmaFatma SSenen BW.inddBW.indd 118181 003-01-113-01-11 15:1715:17 Attending international meetings

2007 TEDDY Open Conference, London, United Kingdom

2008 TEDDY NoE annual meeting, ‘ European symposium on ethics and paediatric Clinical research in Europe’, Marseille, France.

2010 TEDDY Final Open Conference ‘ Paediatric Medicines in Europe. Past, Present and Future’ , Brussels, Belgium.

Teaching

2008-2009 Supervising practicals in Evidence Based Medicine at the Faculty of Medicine, Erasmus MC, Rotterdam

2008-2009 Teaching assistant, practicals in (pharmaco-)epidemiology at the Faculty of Medicine, Erasmus MC, Rotterdam

Other

PhD portfolio 2010 Reviewing for the ‘British Journal of Clinical Pharmacology’, and ‘ European Journal of Clinical Pharmacology’ .

182

EEliflif FatmaFatma SSenen BW.inddBW.indd 118282 003-01-113-01-11 15:1715:17 Dankwoord

EEliflif FatmaFatma SSenen BW.inddBW.indd 118383 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 118484 003-01-113-01-11 15:1715:17 Dankwoord

Na meer dan 3 jaar onderzoek is het dan eindelijk tijd voor het dankwoord! Opgelucht en met veel plezier schrijf ik dit laatste onderdeel van m’n boekje. Iedereen die op wat voor manier dan ook heeft bijgedragen aan de totstandkoming van mijn proefschift wil ik natuurlijk bedanken en een aantal mensen in het bijzonder.

Allereerst wil ik mijn promotor, Miriam Sturkenboom bedanken. Toen ik bijna 4 jaar geleden als nacodeur begon op de afdeling Medische Informatica, had ik nooit durven dromen dat ik een aantal jaar later met mijn eigen proefschrift in m’n handen zou staan! Miriam, heel erg bedankt dat je me die kans hebt gegeven. Jouw begeleiding heb ik al die jaren als zeer plezierig ervaren. Tijdens de wat mindere perioden gaven je enthousiasme en positieve energie mij het nodige zetje om door te gaan.

Vervolgens natuurlijk mijn copromotor, Katia Verhamme. Katia de vele uren samen achter de pc, zwoegend over Foxpro, hebben hun vruchten afgeworpen. Bedankt ook voor je geduld, wanneer ik weer ’s in al mijn ongeduld een syntax of paper had afgeraff eld! Je was een soort ‘moeder op het werk’, zorgzaam en bezorgd om je promovendi.

Ik wil ook graag prof. Den Jongste, prof van den Anker en prof. Ceci hartelijk danken voor hun deelname aan de leescommissie.

Toen ik startte als promovendus, kwam ik in een echte mannenkamer terecht. Gianluca (grote broer), Roelof en Seppe, bedankt voor jullie gezelschap! Ik voelde mij binnen no time thuis op de afdeling. Natuurlijk ook alle andere (oud-)collega’s: Ana, Ann, Carmen, Eva, Inge, Jeanne, Lenonoor, Marissa, Nico, Preci, Rene, Sandra, Silvana, Vera.

Tevens wil ik Desiree, Tineke, Carmen en Sander bedanken voor hun hulp bij de administratieve problemen en Kris, Mees en Marcel voor de soms broodnodige technische hulp.

Mijn vrienden, ‘the uni-gang’, jullie heel erg bedankt voor de vele uren gezelligheid! Sevilay (paranimf), Gülhan, Hatice, Nazli, Shakib, Musa, Mo en Memo; zonder jullie was het de afgelopen jaren toch maar een saaie boel geweest! Ook al hebben de meesten van ons nu andere verplichtingen, ik hoop dat we toch tijd voor elkaar vrij blijven maken.

Mijn familie, pap en mam in het bijzonder, wil ik bedanken voor alle steun en zorg die ik mijn hele leven van ze heb gekregen. De vele weekenden samen zorgden voor de nodige ontspan- ning. Zonder jullie was ik nooit zover gekomen.

185

EEliflif FatmaFatma SSenen BW.inddBW.indd 118585 003-01-113-01-11 15:1715:17 Mijn zus, tevens mijn paranimf, wil ik in het bijzonder bedanken. Mientje, we hebben samen een hele lange weg afgelegd. Je bent er altijd voor me geweest, no matter what, ook nu met mijn promotie. Ik hoop dat ik hetzelfde voor jou kan doen...

Tot slot mijn man Yavuz. Ook al zijn we nog maar net getrouwd, je hebt de afronding van mijn promotie in al zijn hectiek meegemaakt. Hopelijk komen we nu in wat rustiger vaarwater terecht waarin we meer van elkaar gezelschap kunnen genieten. Op naar de rest van ons leven samen... Dankwoord

186

EEliflif FatmaFatma SSenen BW.inddBW.indd 118686 003-01-113-01-11 15:1715:17 List of publi cations

EEliflif FatmaFatma SSenen BW.inddBW.indd 118787 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 118888 003-01-113-01-11 15:1715:17 List of publications

List of Publications

Sen EF, Verhamme KM, Herings R, Neubert A , Felisi M , Sturkenboom MC. THE TEDDY NETWORK: EPIDEMIOLOGICAL TRENDS IN PAEDIATRIC DRUG USE IN EUROPE. European Journal of Hospital Pharmacy Practice, 6 (2007), 22–24.

van der Aa MN, Steyerberg EW, Sen EF, Zwarthoff EC, Kirkels WJ, van der Kwast TH, Essink-Bot ML. Patients’ perceived burden of cystoscopic and urinary surveillance of bladder cancer: a randomized comparison. British Journal of Urology International. 2008 May;101(9):1106-10.

Sturkenboom MC, Verhamme KM, Nicolosi A, Murray ML, Neubert A, Caudri D, Picelli G, Sen EF, Giaquinto C, Cantarutti L, Baiardi P, Felisi MG, Ceci A, Wong IC; TEDDY European Network of Excellence. Drug use in children: cohort study in three European countries. British Medical Journal. 2008 Nov 24;337:a2245.

van der Aa MN, Bekker MD, van der Kwast TH, Essink-Bot ML, Steyerberg EW, Zwarthoff EC, Sen FE, Elzevier HW. Sexual function of patients under surveillance for bladder cancer. British Journal of Urology International. 2009 Jul;104(1):35-40.

Hsia Y, Neubert A, Sturkenboom MC, Murray ML, Verhamme KM, Sen F, Giaquinto C, Ceci A, Wong IC; on behalf of the TEDDY Network of Excellence. Comparison of antiepileptic drug prescribing in children in three European countries. Epilepsia. 2009 Oct 8. [Epub ahead of print]

Sen EF, Verhamme KM, Felisi M, ‘t Jong GW, Giaquinto C, Picelli G, Ceci A, Sturkenboom MC. Eff ects of safety warnings on prescription rates of cough and cold medicines in children below 2 years of age. Accepted for publication in the British Journal of Clinical Pharmacology.

Trifi rò G, Gambassi G, Sen EF, Caputi AP, Bagnardi V, Brea J, Sturkenboom MC. Association of community-acquired pneumonia with antipsychotic drug use in elderly patients: a nested case-control study. Annals of Internal Medicine. 2010 Apr 6;152(7):418-25, W139-40.

Trifi rò G, Dieleman J, Sen EF, Gambassi G, Sturkenboom MC. Risk of ischemic stroke associated with antidepressant use in elderly persons. Journal of Clinical Psychopharmacology. 2010 Jun;30(3):252-8.

Neubert A, Verhamme K, Murray ML, Picelli G, Hsia Y, Sen FE, Giaquinto C, Ceci A, Sturkenboom M, Wong IC; TEDDY Network of Excellence. The prescribing of analgesics and non-steroidal anti-infl ammatory drugs in paediatric primary care in the UK, Italy and the Netlerlands. Pharma- cological Research. 2010 Sep;62(3):243-8.

189

EEliflif FatmaFatma SSenen BW.inddBW.indd 118989 003-01-113-01-11 15:1715:17 Sen EF, Verhamme KM, Neubert A, Hsia Y, Murray M, Felisi M, Giaquinto C, t Jong GW, Picelli G, Baraldi E, Nicolosi A, Ceci A, Wong IC, Sturkenboom MC. Assessment of Pediatric asthma drug use in three European countries; a TEDDY study. European journal of pediatrics. 2010 Sep 2. [Epub ahead of print]. List of publications

190

EEliflif FatmaFatma SSenen BW.inddBW.indd 119090 003-01-113-01-11 15:1715:17 Abo ut the author

EEliflif FatmaFatma SSenen BW.inddBW.indd 119191 003-01-113-01-11 15:1715:17 EEliflif FatmaFatma SSenen BW.inddBW.indd 119292 003-01-113-01-11 15:1715:17 About the author

About the author

Elif Fatma Șen was born on August 5th, 1983 in The Hague, the Netherlands. In 2001 she fi nished her gymnasium education at the Atlas College in Rijswijk. She studied Biomedical Sciences at the University of Amsterdam for one year and started medical school in 2002. In 2007 she obtained her Medical Bachelor and started her PhD research at the department of Medical Informatics of the Erasmus MC. There, she performed the research described in this thesis. In 2008 she obtained her Master of Science degree in Clinical Epidemiology at the National Institute for Health Sciences (NIHES). In 2011 she will start her internship to obtain her medical degree.

193

EEliflif FatmaFatma SSenen BW.inddBW.indd 119393 003-01-113-01-11 15:1715:17